Note: Descriptions are shown in the official language in which they were submitted.
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
STABLE UNBONDED FIBER-REINFORCED ELASTOMERIC
SEISMIC ISOLATORS FOR BASE ISOLATION SYSTEM
FIELD OF THE INVENTION
[0001] This
invention relates to fiber reinforced elastomeric isolator bearings for
use in structural systems including low to medium rise buildings, bridges and
other
structures. In
particular, this invention relates to unbonded bearings which
demonstrate stable rollover deformation.
BACKGROUND OF THE INVENTION
[0002] Studies
of past earthquakes have indicated that major loss of life often
occurs due to collapse of poorly constructed housing. Conventional earthquake-
resistant design of structures has been substantially improved over the last
few
decades. However, in many countries, the engineering services including
design,
construction, and inspection allocated to small and low rise private buildings
and
other structures may not be sufficiently comprehensive to provide an adequate
level of
structural safety against strong or even moderate earthquake events. If the
level of
seismic demand, imposed on these buildings, was reduced through a simple but
reliable base isolation technique, this would result in fewer building
failures and less
loss of life.
[0003] Steel Reinforced
Elastomeric Isolator (SREI) bearings are currently the
most widely used isolators. However, as a result of being heavy and high
priced, their
application has generally been limited to large and expensive structures.
Fiber
Reinforced Elastomeric Isolators (FREIs) employ fibers rather than steel
plates, as the
reinforcement sheet. FRET bearings can provide adequate levels of vertical and
horizontal stiffnesses as required in a base isolation device. Furthermore,
unique
aspects such as potentially low manufacturing cost, higher energy dissipation
capability, light-weight, and, the possibility of being produced in long
rectangular
strips and being cut to size as required, provides promising advantages for
this type of
bearing.
[0004] Kelly (1999, Analysis of Fiber-Reinforced Elastomeric Isolators,
Journal of
Seismology and Earthquake Engineering (JSEE), 2 (1), 19-34) conducted an
1
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
experimental study on cylindrical handmade bearings consisting of high damped
rubber reinforced with Kevlar (TM) fibers. From the test results, it was
revealed that
fiber reinforcement can provide acceptable compression stiffness.
Additionally, lack
of flexural rigidity of reinforcement was shown to have a small effect on the
horizontal stiffness of the bearing. The generated hysteresis loops under
combined
compression and shear showed the same general characteristics as a traditional
SREI
bearing with a stable behavior up to a peak shear strain of 150%. Furthermore,
damping ratios higher than anticipated were obtained which revealed a new
source of
energy dissipation. This was an unexpected advantage of using fiber as
reinforcement
in elastomeric bearings.
[0005] Seven
rectangular carbon-FREIs were tested by Kelly (2002, Seismic
Isolation Systems for Developing Countries, Earthquake Spectra, 18 (3), 385-
406)
under both a compression load to measure the compression stiffness and a
combination of compression and shear loading to measure the horizontal
stiffness and
effective damping. For the latter case, the test was repeated for orientations
of 0, 90,
and 45 degrees with respect to the longitudinal direction of the strip. It was
observed
that loading along 0 degree produces stiffening in the hysteresis loops,
whereas along
90 degrees (i.e., the cross direction) softening behavior tends to occur.
Loading along
45 degree produced neither softening nor stiffening. Experimental results
confirmed
that it is possible to produce a strip FREI that matches the behavior of a
SREI. The
measured horizontal stiffnesses and the maximum accommodated displacement
indicated that the concept of carbon strip-FREI is viable.
[0006] Moon et al. (2003,
Mechanical Property Analysis and Design of Shock
Absorber System Using Fiber Bearing by Experimental Method, JSME International
Journal, 46 (1), 289-296) compared the performance of a cylindrical carbon-
FREI to
that of the same size SREI. The difference between the steel plate and the
fiber
thickness was adjusted by using more layers of fiber and rubber in the FREI.
Accordingly, bulging of the FREI was smaller than that of the SREI due to the
thinner
layers of rubber in the FREI. Unlike previous studies by Kelly where the
bearings
were built without end plates and were not bonded to the test machine during
the test,
both bearings were bonded to thick end plates. The researchers concluded that
the
performance of the FREI is superior to that of the SREI. However, due to
insufficient
2
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
information regarding the details of the tested isolators, the basis of
comparison is not
clearly identified. Accordingly, comparative studies between FREIs and SREIs
still
need to be conducted.
[0007] Summers et
al. (2004, Development of new base isolation devices for
application at refineries and petrochemical facilities, 13th World Conference
on
Earthquake Engineering Vancouver, B.C., Canada, August 1-6, Paper No. 1036)
conducted an experimental study on prototype rectangular carbon FREIs as a
potential
seismic protection strategy for liquid storage tanks. The bearings consisted
of high
damped rubber compound and were subjected to a maximum 100% shear deformation
under constant vertical compression. The resulting hysteresis loops showed
stable
behavior.
[0008] FREI
bearing or isolation pads are used in both new and retrofit
construction to provide vibration damping for residential and commercial
properties
and prevent structural collapse during seismic events. Various FRET bearings
are
disclosed in prior art issued patents and published patent applications.
[0009] US
Patent Publication No. 2004/0123530 describes a system for protecting
a structure from seismic ground motion comprising a horizontal bearing surface
and
damping elements consisting of unreinforced rubber slabs and fiber-reinforced
elastomeric mats. Each mat consists of a piece of a rubber mat reinforced with
fibers
extending parallel to the surfaces of the mat. The horizontal bearing surface
is covered
with several layers of the damping elements.
[0010] US
Patent No. 5,014,474 describes an apparatus for limiting the effect of
vibrations between a structure and its foundation having two types of
elastomeric load
bearing pads. One pad absorbs vertically and horizontally applied forces. The
other
pad can accommodate sliding motion between the structure and the foundation
and
can absorb horizontally applied forces.
[0011] US
Patent No. 5,904,010 describes a fiber reinforced elastomeric seismic
isolation bearing. The bearing is a laminate block of material with
alternating layers
of elastomer and woven fiber mesh having fibers 0.1 to 1 mm made of graphite
or a
3
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
polyamide polymer, such as Kevlar (TM). This patent however describes bearings
with vulcanized layers and pretensioned fibers.
[0012] US
Patent No. 4,695,169 illustrates a structural bearing made with an
elastomeric core imbedded with a matrix of twisted Kevlar (TM) cord. Rollover
is
not disclosed in this patent.
[0013] US
Patent No. 4,593,502 describes a rectangular energy absorber
constructed from layers of resilient materials and stiffening materials. The
energy
absorber also includes a steel spring and a lead core. The absorber can also
be
constructed with alternating layers of resilient and stiffener material. The
layers
described in this patent are not vulcanized and the fibers are not
pretensioned.
Rollover performance is not disclosed in this patent.
[0014] US Patent No.
4,887,788 describes the general state of the art as a base
isolation pad to absorb energy. The disclosed device is made of an elastomeric
material with a flexible reinforced tubular restraining shell.
[0015] US
Patent No. 5,233,800 describes an earthquake proofing device designed
to dampen the effects of seismic activity. The pads of this device incorporate
laminated layers of synthetic rubber with specific characteristics and rigid
plates made
from wire.
[0016] US
Patent No. 5,438,806 describes a composition for designing a device for
vibration damping of released energy. The device consists of a number of
configurations of elastomeric layers containing air or fluid voids.
[0017] The bearings of the prior art consider the rollover effect as an
undesirable
parameter which decreases stability.
[0018] There remains a need to provide a base isolation system which is
intended
for cost effective seismic mitigation of structures such as small and medium
low-rise
buildings and bridges. Such a system would require FRET bearings in an
unbonded
application, which demonstrate increased stability with stable rollover
deformation.
4
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
SUMMARY OF THE INVENTION
[0019] The
present invention addresses the above drawbacks by providing a cost-
effective base isolation system for seismic mitigation of structures such as
small and
medium low-rise buildings and bridges. Such a system utilizes FREI bearings in
an
unbonded application, while demonstrating increased stability with stable
rollover
deformation.
[0020] In one
aspect of the present invention, there is provided an elastomeric
seismic isolator bearing comprising a plurality of elastomeric layers
reinforced with a
plurality of fiber layers, wherein the bearing is capable of exhibiting stable
rollover
when subjected to a lateral load in an unbonded application.
[0021] The
elastomeric layers comprise one or more filled or unfilled elastomers;,
with the elastomer preferably having a hardness of between 30- 60 Durometer
Shore
A; and preferably have a damping of between 2% to 20%. The elastomer is
preferably selected from the group consisting of, natural gum rubber,
neoprene, butyl
rubber and urethanes.
[0022] The plurality
of fiber layers comprise a bi-directional fabric, a tri-
directional fabric or alternating unidirectional fabrics with perpendicular
orientation.
In addition, the fiber layers comprise woven carbon fiber, nylon, Kevlar TM (a
para-
aramid synthetic fiber) or aramid fiber. The fiber layers have a fiber to
matrix volume
fraction of preferably between 30% to 70%.
[0023] A hot
vulcanization process , cold vulcanization process, adhesive or other
mechanical bonding can be used to bond one or more of the elastomeric layers
to one
or more of the fiber layers. The bearing has a shape factor of preferably
between 5 to
25, and an aspect ratio of at least 2. In addition, the bearing has a fiber to
elastomer
volume fraction of at least 5%.
[0024] The bearing can have one or more modified vertical faces, or have one
or
more vertical faces fitted with a preformed attachment. Alternatively, contact
surfaces
above and below the bearing can each have a modified shape.
5
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[0025] For seismic mitigation of structures, the bearing of the present
invention
has a seismic isolation period of at least 1 second.
[0026] In another aspect of the present invention, there is provided an
elastomeric seismic isolator bearing comprising a plurality of natural gum
rubber
layers interspersed with a plurality of bi-directional carbon fiber layers,
wherein said
bearing is utilized in an unbonded application and is capable of exhibiting
stable
rollover when subjected to a lateral load.
[0027] One or more of the natural gum rubber layers is bonded to one or more
of
the bi-directional carbon fiber layers using mechanical bonding. The bearing
has a
shape factor of preferably between 5 to 25; an aspect ratio, also known as a
second
shape factor, of at least 2; and a fiber to elastomer volume fraction of at
least 5%. In
addition, the bearing has a seismic isolation period of at least 1 second. The
elastomeric seismic isolator bearing is used for seismic mitigation of a
structure with
a fixed base period of between 0.1 to 0.5 seconds.
[0028] In yet another aspect of the present invention, there is provided
an
elastomeric seismic isolator bearing comprising a plurality of neoprene layers
interspersed with a plurality of bi-directional carbon fiber layers, wherein
said bearing
is utilized in an unbonded application and is capable of exhibiting stable
rollover
when subjected to a lateral load.
[0029] One or more of the neoprene layers is bonded to one or more of the bi-
directional carbon fiber layers using an adhesive or other mechanic bonding.
The
bearing has a shape factor of preferably between 5 to 25; an aspect ratio of
at least 2;
and a fiber to elastomer volume fraction of at least 5%. In addition, the
bearing has a
seismic isolation period of at least 1 second. The elastomeric seismic
isolator bearing
is used for seismic mitigation of a structure with a fixed base period of
between 0.1 to
0.5 seconds.
[0030] In yet
one more aspect of the present invention, there is provided a method
for seismic mitigation of a structure comprising: placing a plurality of
elastomeric
6
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
seismic isolator bearings between the structure and a foundation of the
structure; the
bearing comprising a plurality of elastomeric layers reinforced with a
plurality of fiber
layers, wherein the bearing is not bonded to the structure and is capable of
exhibiting
stable rollover when subjected to a lateral load.
[0031] The
elastomeric seismic isolator bearing has a seismic isolation period of
at least 1 second and the structure has a fixed base period of between 0.1 to
0.5
seconds.
[0032] The
elastomeric layers comprise one or more unfilled or filled elastomers;,
with the elastomer preferably having a hardness of between 30- 60 Durometer
Shore
A; and preferably have a damping of between 2% to 20%. The elastomer is
preferably
selected from the group consisting of natural gum rubber, neoprene, butyl
rubber and
urethanes.
[0033] The
plurality of fiber layers comprise a bi-directional fabric, a tri-
directional fabric or alternating unidirectional fabrics with perpendicular
orientation.
In addition, the fiber layers comprise woven carbon fiber , nylon, para-aramid
synthetic fiber or aramid fiber. The fiber layers have a fiber to matrix
volume fraction
of preferably between 30% to 70%.
[0034] A hot
vulcanization process, cold vulcanization process, adhesive or other
mechanical bonding.can be used to bond one or more of the elastomeric layers
bonded
to one or more of the fiber layers. The bearing has a shape factor of
preferably
between 5 to 25, and an aspect ratio of at least 2. In addition, the bearing
has a fiber
to elastomer volume fraction of at least 5%.
[0035] The bearing can have one or more modified vertical faces, or have one
or
more vertical faces fitted with a preformed attachment. Alternatively, contact
surfaces
above and below the bearing can each have a modified shape.
[0036] While the present invention is suitable for seismic mitigation,
it is also
applicable for non-seismic events. In addition, a worker skilled in the art
would
7
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
appreciate that all elastomers that demonstrate stable rollover deformation in
the
described conditions and which are suitable for use as bearings may be used in
the
present invention.
BRIEF DESCRIPTION OF THE DRAWINGS
[0037] Fig. 1
shows the test setup and instrumentation arrangement used in one
embodiment of the invention. Fig la is a side view. Figs. lb and lc are
sectional
views 1-1 and 2-2, respectively, from Fig. la.
[0038] Fig. 2
shows the vertical load-deflection relationship under 64 kN vertical
load for two bearings B1 and B2 in an embodiment of the present invention.
[0039] Fig. 3 shows
the lateral load-displacement behavior for Bearing B1 (aspect
ratio RBI = 1.9, and tr, BI = 94 mm), in a series of cyclic tests having
amplitudes of 25,
50, 75, 100, and 125% tr, al, under constant vertical compression (a) 1.6 MPa,
(b) 3.2
Mpa.
[0040] Fig. 4 shows
the lateral load-displacement behavior, under constant 1.6
MPa vertical compression, in a series of cyclic tests having amplitudes 25,
50, 75,
100, 125, 150, and 200% tr Fig. 4a shows the bearing NB1 (aspect ratio RNBI =
2.5,
and tr, NB! = 70.5 mm), and Fig. 4b shows the bearing NB2 (aspect ratio RNB2=
2.9, tr;
NB2 = 61.1 mm).
[0041] Fig. 5
shows an embodiment of a base isolated masonry building (all
dimensions in meter). Fig. 5a shows an elevated view and Fig. 5b shows a
sectional
view along A-A of Fig. 5a. Solid rectangles in Fig. 5a and solid squares in
Fig. 5b
represent the carbon-FREI bearings.
[0042] Fig. 6
depicts a mass-spring-dashpot idealization used to model the base
isolated building of one embodiment of the present invention.
[0043] Fig. 7
shows hysteresis loops for carbon-FRET bearings with 5% damping
ratio.
8
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[0044] Fig. 8 shows the response time histories of a base isolated
building with 5%
damped bearings. Fig. 8a depicts the lateral displacement at isolators level
and Fig.
8b depicts the isolated building period.
[0045] Fig. 9 shows the absolute acceleration of the fixed base (FB) and
base
isolated (BI) buildings. Fig. 9a depicts accelerations at the first floor and
Fig. 9b
depicts accelerations at the roof.
[0046] Fig. 10 shows the response history of the FB and BI buildings. Fig.
10a
depicts the base shear and Fig. 10b depicts the inter-storey drift at 2'
storey.
[0047] Fig. 11 shows a sketch of seismic isolator sheets B1 & B2.
[0048] Fig. 12 shows hysteresis loops corresponding to the final cycle of
lateral
displacement at each displacement amplitude (lateral displ. ranges: 25%, 50%,
75%,
100%, 150%, and 200%; tr; displ. rate = 30 mm/s). Fig. 12a is of Bearing B2-1
and
shows parallel response, and Fig 12b is of Bearing B1-1 and shows diagonal
response.
[0049] Fig. 13 shows the influence of displacement rates lower than 114
mm/s on
lateral response at 100% tr (19 mm) displacement amplitude. Values represent
the
average reduction in all 3 displacement cycles. Fig. 13a shows a reduction in
effective
stiffness and Fig 13b shows a reduction in effective damping.
[0050] Fig. 14. shows the parallel lateral response (00) of Bearings B1-3
and B2-3
(lateral displ. ranges: 25%, 50%, 75%, 100%, 150%, and 200% tõ displ. rate =
76
mm/s). Fig. 14a is for input signal at 6 ascending displacement amplitudes,
Fig 14b
shows hysteresis loops of Bearing B1-3 for all cycles, Fig 14c is for input
signal at 6
descending displacement amplitudes, and Fig 14d shows hysteresis loops of
Bearing
B2-3 for all cycles.
9
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[0051] Fig. 15
shows photographs taken from parallel (00) response of Bearing
B1-3 at different amplitudes of lateral displacement ; (a) 0% tr (0 mm), (b)
100% tr
(19 mm), (c) 150% tr (28.5 mm), (d) 200% tr (38 mm).
[0052] Fig. 16 shows the
lateral response values for Bearings B1-3 (at 0 ), B2-1
(at 900), and B2-3 (at 900). Response values represent the average of all 3
cycles at
each displacement amplitude. Fig 16a shows the effective stiffness and Fig 16b
shows
the effective damping.
[0053] Fig. 17 shows photographs taken of diagonal (45 ) response of
Bearing
B1-3 at different amplitudes of lateral displacement; (a) 0% t, (0 mm), (b)
100% t, (19
mm), (c) 150% t, (28.5 mm), (d) 200% tr (38 mm).
[0054] Fig. 18
shows the lateral response characteristics of Bearings B1-3 at 45
orientation. Response values in Fig 10(b) represent the average values at each
displacement amplitude. Fig 18a show hysteresis loops for all cycles and Fig.
18b
shows response variations with displacement amplitude.
[0055] Fig. 19
depicts the lateral response properties of the bearings in lateral
cyclic tests with different patterns of displacement amplitude. Values are the
average
of all cycles at each amplitude level. Fig. 19a shows an ascending amplitude
pattern
and Fig. 19b shows a descending amplitude pattern.
[0056] Fig. 20
shows the influence of 50% variation in the 1.6 MPa vertical
design pressure on lateral parallel response of Bearings B1-3, B2-1, and B2-3
(lateral
displacement amplitude = 100% t, or 19 mm, displ. rate = 76 mm/s). Fig. 20a
depicts
the variation in effective horizontal stiffness and Fig. 20b depicts the
variation in
effective damping.
[0057] Fig. 21 shows
the time history of the input signal utilized for the vertical
tests on the bearings of the present invention.
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[0058] Fig. 22
shows the vertical compressive stress-strain behavior of Bearing
B2-3 under 0.8, 1.6, and 2.4 MPa vertical pressure. Fig 22a shows the stress-
strain
plot and Fig. 22b shows the effective compressive modulus representation.
[0059] Fig. 23 shows the lateral response characteristics of bearings in
the repeated
cyclic testing. Response values are the average of all cycles at each level of
lateral
displacement amplitude. Fig 23a shows the effective horizontal stiffness and
Fig. 23b
shows the effective damping.
[0060] Fig. 24 shows hysteresis loops for repeated cyclic testing of
Bearing B1-3.
Fig 24a is at 00 Response and Fig. 24b is at 900 Response.
[0061] Fig. 25
shows retest versus virgin response of Bearing B1-3 under cyclic
lateral displacements at 00. Fig 25 a shows the half-cycle stable (scragged)
loops of
the virgin bearing versus the retested bearing, for lateral displacement
ranges: 100%,
150%, and 200%t1, and Fig. 25b shows the average response variations in the
retested
bearing with respect to the virgin bearing.
[0062] Fig. 26
shows the longitudinal lateral response of the virgin Bearing TB1
under constant 1.6 MPa vertical pressure at lateral displacement ranges: 25%,
50%,
75%, 100%, 150%, and 200% tr and displacement rate = 76 mm/s.
[0063] Fig. 27
shows photographs of Bearing TB1 taken during lateral cyclic
testing. No bonding finishing coats were applied to the perimeter surfaces of
this
bearing. Fig 27 a depicts the Start (0% tr), Fig. 27b is at 200% tr (38 mm) -
cycle 3,
Fig 27c is at 250% tr (47.5 mm) - cycle 3, Fib. 27d is at 300% tr (57 mm) -
cycle 1,
Fig. 27e is at 300% tr (57 mm) - cycle 3, and Fig. 27f is at the End (0% tr).
[0064] Fig. 28
shows the longitudinal lateral response for Bearing TB1 (p = 1.6
MPa, lateral displ. ranges: 100%, 200%, 250%, and 300% tr, displ. rate = 76
mm/s).
Fig 28a shows the input lateral displacements and Fig 28b shows the hysteresis
loops
for the ri cycle of each amplitude.
11
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[0065] Figs.
29a-g show modifications to the vertical faces of the bearing and
contact surfaces.
DETAILED DESCRIPTION OF THE EMBODIMENTS
[0066] The base
isolation system of the present invention utilizes FREI bearings
that are located between the superstructure and substructure with no bonding
at the
contact surfaces of the bearings. This is denoted as an "unbonded
application". In
such an application, FREI bearings, with the appropriate aspect ratio, exhibit
stable
rollover deformation. In a FREI bearing utilized in unbonded application as
the base
is displaced laterally, the top and bottom faces of FREI-bearing roll off the
upper and
lower contact supports, respectively. This lateral deformation is denoted as
"rollover
deformation". It occurs as a result of both unbonded application and lack of
flexural
rigidity of fiber reinforcement sheets. A properly designed FREI bearing can
sustain
very large lateral displacements while remaining stable, by maintaining a
positive
incremental lateral load resisting-capacity throughout the hysteresis loops.
As such,
the resulting permissible deformation is called "stable rollover" (SR)
deformation.
SR-deformation results in a significant decrease in the horizontal stiffness
of the
bearing and adds to its efficiency as a seismic isolator. An unbonded FREI
bearing
that exhibits SR-deformation is denoted as a "stable unbonded" (SU) FREI. In a
SU-
FREI bearing, the "full contact vertical face lateral displacement" (Ok)
occurs when
the original vertical faces of the bearing completely come in contact the
horizontal
upper and lower supports. As a result, significant stiffening in the
hysteresis loops is
observed which effectively places a limit on the extreme lateral displacements
that
can occur under unanticipated seismic excitation levels. Rollout instability
lateral
displacement, denoted as 8max, for this particular type of bearing is
significantly larger
than 5f, .
[0067] Square carbon-FREI bearings with different aspect ratios, in one
embodiment, were tested under compression and cyclic shear to evaluate the
mechanical characteristics of the bearings employed in one final design. An
analytical
approach was used to model the lateral load-displacement hysteresis of the
carbon-
FREI bearings. The dynamic response of a representative Base Isolated (BI)
building
as well as its corresponding Fixed Base (FB) model against an input earthquake
12
CA 02727039 2014-09-26
confirmed the seismic mitigation efficiency of the base isolation system. The
BI-
system in the above mentioned BI-building comprised SU-FREIs that were located
relatively close to one another. As a result of the short distance between the
adjacent
bearings, significant cost savings are achieved in both the concrete tie-beam
located at
the top of the bearings to carry the superstructure between them, and the
foundation
system.
[0068] In a fiber reinforced elastomeric isolator (FREI) bearing under
vertical
compression loads, elastomer layers tend to extend laterally. However, if a
strong
bond exists between the elastomer and the fiber reinforcement, this lateral
extension
in the elastomer is restrained by the reinforcement layers. Accordingly, fiber
reinforcement layers are stretched in tension and the elastomer layers
experience
compression except at the extreme outer edges where the bulging elastomer is
subjected to tension. Fiber reinforcement sheets are made of either bi-
directional
fabric or alternating unidirectional fabrics with perpendicular orientation.
[0069] The design properties of unbonded square carbon-FREI bearings,
including
displacement characteristics and damping properties were evaluated through
compression (vertical) and cyclic shear (horizontal) tests. The relevant
design
provisions of ASCE 7 (2005), a commonly employed standard in the design of
seismically isolated structures, were utilized to verify the lateral response
of the tested
bearings.
Example 1
[0070] In one embodiment of the present invention, two carbon-FREI
bearings
were built. Each bearing consisted of bonded layers of intermediate elastomer
and
reinforcement as well as two bonded thinner elastomeric covers at the top and
bottom.
The total thickness of rubber layers was tr = 94 mm. The width was b = 200 mm
and
the total height was approximately h = 105mm. The shape factor, defined as the
ratio
of loaded area to load free area of the elastomer layer, was S = 10.6 and the
aspect
ratio (i.e., width to total height ratio, sometimes called the second shape
factor) of the
bearings was approximately R = 1.9. The elastomer was a soft compound of
natural
13
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
gum rubber with a hardness of 40 Durometer Shore A and bi-directional carbon
fabric was utilized as the reinforcement. The bearings were made without the
use of a
mold and all vertical faces were trimmed with a band saw. A finishing coat of
bonding compound was applied to the vertical faces of each bearing to prevent
any
premature delamination.
[0071] Fig. la
shows a sketch of the test setup which was designed to perform
both the vertical and horizontal (cyclic) testing on the bearings (40). The
bearing (40)
was located between two 100 mm thick steel platens (25) and (50). Horizontal
displacements were imposed via a horizontal hydraulic actuator (45) attached
to the
lower platen (50). The lower platen (50) was mounted on four linear bearings
(65) and
could move 150 mm with respect to the upper platen (25). The upper platen
(25) was
braced to the body of the horizontal actuator (45) at its lateral sides
through two arms
(30) providing a self-equilibrating system. The entire apparatus was mounted
on a
stand (75). The bearing was subjected to near pure shear as the action and the
reaction
horizontal forces passed through its midheight.
[0072] Figs. lb
and lc, which are sectional views 1-1 and 2-2 shown in Fig. la,
illustrate the instrumentation used for the test setup. Load cell LC#1 (60)
was used to
measure the horizontal loads corresponding to different horizontal deflections
imposed on the bearing (40). The relative horizontal movements between the
lower
and upper platens (50, 25) were monitored through a string pot (35 in Fig. 1
a). Four
laser displacement transducers (LDT), (80 81 82 and 83),) were used to measure
the
vertical deflection of the bearing (40) at its four sides during the vertical
testing. The
vertical deflection was calculated by taking the average value of the LDTs
readings.
To achieve a symmetrical setup, four identical load cells (61, 62, 63 and 64)
were
used to measure the vertical load. Steel bearings (10) were placed between the
upper
platen (25) and these four load cells (61, 62, 62 and 64) to ensure minimum
horizontal
load would be transferred to the load cells measuring the vertical load. The
vertical
load was applied through a vertical ram (5).
[0073] The vertical compression modulus as well as the maximum vertical
deflection of the bearings at 1.6 MPa design vertical pressure (64 kN), were
evaluated
through vertical testing. Figs. 2a and 2b present the results of the vertical
test on
14
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
Bearings B1 and B2, respectively. The slope of the dashed straight line
represents the
average stiffness of each bearing during the cyclic reversals of loading.
[0074] From the
test results, the vertical stiffness of the bearings were calculated
as; (Kv) B1 = 76.13 kN/mm, and (Kv) B2 = 72.87 kN/mm. These correspond to
effective compression modulus values of (Er) BI = 178.9 MPa, (Ec) B2 = 171.2
MPa,
and effective vertical frequency values of (fõ) Bi = 17.2 Hz, and (fv) B2 =
16.8 Hz,
respectively, which are considered satisfactory values for isolation
applications.
Accordingly, the bearings of this embodiment provide sufficient vertical
stiffness for
vertical pressures around the design value of 1.6 MPa.
[0075] Cyclic
tests were performed under horizontal displacement control and
vertical load control. The following test procedure was used. While the
bearing was
subjected to a constant vertical pressure (1.6 MPa), four fully reversed
sinusoidal
cycles of horizontal displacements having amplitudes of 25% tõ 50% tõ 75% tõ
100%
tr, 125% tr, 150% tõ and 200% tr (where tr is the total thickness of rubber
layers in
each bearing) were applied. For each amplitude, the bearing was vertically
loaded up
to the target load. Subsequently, four cycles of horizontal displacement were
applied
after which the bearing was vertically unloaded.
[0076] The
horizontal load-displacement behavior of Bearings B1 and B2 at
different values of peak horizontal displacement was found to be similar. Fig.
3a
shows the horizontal load-displacement curves of Bearing B1 subjected to 25,
50,
100, and 125% tr horizontal displacement. As illustrated, the bearing behaved
linearly
up to approximately 75% tr horizontal displacement with an effective
horizontal
stiffness Kh,Bi= 0.13 kN/mm and 2.1% effective damping ratio calculated using
equations provided by ASCE 7 (2005, American Society of Civil Engineers.
"Minimum design loads for buildings and other structures." ASCE/SEI 7-05). At
larger lateral displacements, nonlinear response with a tendency of softening
can be
observed. This softening was associated with the unstable rollover deformation
of the
bearing.
[0077] To
evaluate the influence of higher vertical pressure on improving the
horizontal response of the bearing, cyclic tests were duplicated under a
larger constant
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
vertical pressure of 3.2 MPa (see Fig. 3b). At 75% tr horizontal displacement,
the
effective horizontal stiffness of the bearing decreased to 0.12 kN/mm and the
effective damping ratio increased to 5.2%. A significant increase in the
damping
ratios of FREIs due to higher vertical pressure was a unique feature not found
in
SREIs. This confirms one of the advantages of using fibers as the
reinforcement. At
lateral displacements larger than 75% tr, as indicated in Fig. 3b the bearing
performed
nonlinearly and again tended to show a softening type behavior even though a
higher
vertical pressure (twice as larger as the design value) was applied.
[0078] No damage was visible after the cyclic tests were completed.
Furthermore,
the repeatable pattern of the load hysteresis loops, including the maximum
lateral
displacements, demonstrated that negligible damage occurred to the bearings.
As a
result of the softening behavior, the performance of Bearings B1 and B2 with
an
aspect ratio of R = 1.9 was considered unacceptable. In an attempt to improve
the
lateral load-displacement behavior of the bearings, the aspect ratio was
increased by
reducing the bearing thickness. In order to evaluate the influence of
different aspect
ratios on improving the lateral response of the bearings, new Bearings NB1 and
NB2
were cut from B1 and B2 to achieve aspect ratios of RNBI = 2.5 and RNB2 = 2.9,
respectively.
[0079] Fig. 4 shows the hysteresis loops of NB1 when subjected to cyclic
lateral
displacements with different amplitudes ranging from 25% to 200% t
-r,NB1 (where total
thickness of rubber layers tr, NB1= 70.5 mm).
[0080] As illustrated in Fig. 4a, Bearing NB1 of this embodiment performed
almost linearly up to 75% tr, NB1":-..- 53 mm lateral displacement with an
effective
stiffness Kb, NB1= 0.21 kN/mm and effective damping ratio of 1.5%. However, at
larger lateral displacements, due to the rollover deformation of the bearing,
the
tangent horizontal stiffness dramatically dropped to its minimum-yet-positive
value.
At extreme lateral displacements (larger than 150% tr, NBi) when the vertical
faces of
the bearing contacted the upper and lower platens, a hardening behavior was
observed. This was considered to be advantageous as it imposed a limit on the
maximum lateral displacement of the bearing and guarantees stability of the
device
against the maximum considered earthquake.
16
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[0081] Bearing
NB2, with a total thickness of 69 mm, was cut through the
thickness of Bearing B2 so that an aspect ratio of RNB2 = 2.9 was achieved. As
a
result, the total thickness of rubber layers of the bearing was reduced to r
NB2 = 61.1
mm.
[0082] Fig. 4b
shows the hysteresis loops from a set of cyclic tests conducted on
Bearing NB2. This bearing showed a linear behavior up to approximately 75
= - - - t
r, NB2
46 mm lateral displacement with an effective stiffness Kh, NB2 = 0.25 kNimm
and an
effective damping ratio of 1.3%. However, due to rollover deformation at
larger
lateral displacements, the tangent horizontal stiffness of the bearing
gradually
decreased to its minimum value. Analogous to Bearing NB1, at extreme lateral
displacements when the vertical faces of the bearing contacted the upper and
lower
platens, a hardening behavior was observed.
[0083] All
bearings of this embodiment exhibited a rollover deformation when
subjected to lateral loads. This rollover deformation resulted from the
unbonded
boundary condition of the bearing at its contact surfaces as well as the lack
of flexural
rigidity in the fiber reinforcement layers. This deformation, which caused a
significant
reduction in the horizontal stiffness of the bearing, is considered acceptable
if the
resulting tangent horizontal stiffness of the bearing remains positive. In
such a case,
when the vertical faces of the bearing touched the upper and lower platens at
the
extreme lateral displacements, a hardening behavior was observed. This
hardening
placed a limit on the maximum lateral displacement of the bearing and ensured
its
overall stability.
[0084] As previously indicated, such an admissible deformation is
denoted as a
"stable rollover" or SR deformation. A stable rollover deformation neither
results in a
softening behavior, nor does it compromise the overall stability of the
bearing.
Furthermore, being a seismic isolation device, it adds to the efficiency of
the bearing
as the horizontal stiffness is reduced. As a result, this unique performance
is
considered advantageous for application of FREIs with unbonded contact
surfaces.
17
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[0085] Considering all the bearings tested, the lateral load-displacement
performance of Bearings B1 and B2 was determined to be not acceptable as they
exhibited softening and negative tangent horizontal stiffness due to unstable
rollover
deformation (Fig. 3a). The same unacceptable behavior was observed when the
vertical compression was doubled (Fig. 3b). It was found that, for a FREI with
the
same material properties and shape factor (S), the aspect ratio (R) of the
bearing
played an important role in achieving a stable rollover deformation.
[0086] Similar
to SREI behavior, an increase in the vertical pressure reduced the
horizontal stiffness of a FREI. However, a higher vertical pressure resulted
in an
increase in the energy dissipated by a FREI (see Figs. 3a and 3b). This was
attributed
to an increase in the in-plane tension in the fibers. The carbon fiber fabric
of this
embodiment consisted of strands of fibers woven together. When a FREI bearing
was
deformed laterally, the reinforcement layers were curved due to a lack of
flexural
rigidity. Accordingly, the in-plane tension in the curved fibers forced the
strands to
slip against each other dissipating energy through friction.
[0087] As
illustrated in Fig. 4a, the nonlinear behavior of Bearing NB1, due to a
stable rollover deformation, had three distinct response components. For minor
earthquakes, the bearing can be designed to remain within its initial stage of
lateral
load-displacement response with an effective horizontal stiffness Khl,NB1=
0.21
kN/mm. To achieve the maximum efficiency against Design Basis Earthquakes
(DBE), the effective horizontal stiffness dropped to the minimum value Kh2,
NB1= 0.13
kN/mm. At Maximum Considered Earthquake (MCE), with 2% probability of
exceedence within 50 years (2,500 year return period), the bearing was allowed
to
undergo larger lateral displacements and exhibited a hardening behavior. The
effective horizontal stiffness at the latter case was Kh3, NB1= 0.15 kN/mm.
[0088] An
initial horizontal stiffness of 0.21 kN/mm, in a bearing subjected to a
constant vertical compression of 1.6 MPa, resulted in an isolated period of
1.1 S.
However, depending on the magnitude of earthquake energy, the effective
horizontal
stiffness could drop to as low as 0.13 kN/mm and the isolated period could
reach a
value of 1.4 s which resulted in increased seismic mitigation.
18
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[0089] At DBE, the design lateral displacement of Bearing NB1, can vary
between
100% to 150% tr,NB1 (70 mm to 105 mm). Test results have shown that NB1 can
sustain large cyclic lateral displacements up to a maximum value 140 mm (200%
tr,NB1) with no significant stiffness degradation. Furthermore, after the
completion of
the cyclic tests no delamination or any other damage was observed in the
bearing.
Accordingly, at MCE, an upper limit of 140 mm is acceptable for design
purposes as
the maximum admissible lateral displacement of this bearing.
[0090] Comparing Figs. 4b and 4a, the nonlinear behavior of Bearing NB2 was
not as significant as NB1. The effective horizontal stiffnesses of NB2
corresponding
to 75, 125, 150, and 200% t
.r, NB2 lateral displacements were calculated to be 0.25,
0.20, 0.18, and 0.18 IcNimm, respectively. Therefore, the isolated period of
Bearing
NB2 fluctuated between 1 s to 1.2 s. In contrast to NB1, the effective
horizontal
stiffness of NB2 gradually decreased with an increase in the amplitude of
lateral
displacement.
[0091] Due to
the smaller total volume of carbon fiber reinforcement utilized in
NB1, the bearing of this embodiment exhibited lower damping ratios than
Bearing
BI. Depending on the amplitude of lateral displacement, effective damping of
Bearings NB1 varies from 1% to 3% which is less than the required value. The
smaller volume of carbon fiber reinforcement on NB2, compared to that of NB1,
implied lower damping ratios of Bearing NB2. Test results confirmed this
expectation. Depending on the amplitude of lateral displacement, the effective
damping of Bearing NB2 varied from 0.8% to 2.3%.
[0092] After
completion of the cyclic tests, no damage in NB2 was visible.
Additionally, the stable pattern and repeatability of hysteresis loops at
different
displacement amplitudes, including 122 mm (200% t
-r, NB2), indicated that no
noticeable damage had occurred inside the bearing. As a result, it was found
the
bearing can be safely subjected to the maximum 122 mm lateral displacement.
[0093] The
efficiency index of a bearing with nonlinear lateral load-displacement
behavior can be defined as the ratio of maximum instant isolated period to the
19
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
minimum achievable isolated period. This ratio for Bearings NB1 and NB2 were
1.27
and 1.20, respectively. Therefore, NB1 provided more seismic mitigation than
NB2.
[0094] As a
result, for a FREI bearing with the given width, shape factor, vertical
load, and elastomer modulus, by providing higher efficiency, an aspect ratio
of 2.5
was determined to be close to the optimum value for this particular
embodiment.
[0095] The
following 2-storey masonry building (110) (shown in Fig. 5a), located
in a region with high seismic risk, is a representative superstructure for use
with
embodiments of the present invention to be seismically isolated. The building,
with
square base dimensions of 5 m by 5 m, shown as (120) in Fig. 5b), was
constructed
with 20 cm, 15 MPa hollow concrete masonry blocks (125). As shown in Fig. 5b,
the
length (130) of each block (125) is 1.40 m; the distance between each block
(135) is
1.10 m; and while the space (140) between each corner and an end block is 0.55
m..
The height (145) of each floor is 2.8 m, while the thickness (150) between
floors is
0.15 m. To achieve a simple and well defined masonry structural system, each
shear
wall was considered to have a rectangular cross section with no opening
inside, and
was assumed to be fully grouted. Intentionally, there were no intersections
between
the perpendicular walls. Although this structural system may be somewhat more
simple than a real building, its dynamic characteristics were in the same
range as a
typical masonry shear wall structure.
[0096] The base
isolation system of this embodiment, as indicated in Fig. 5b,
consisted of 8 square carbon FREI bearings (90). Three of these bearings (90)
are
shown in Fig. 5a. An unbonded application for the bearings was utilized which
means,
unlike the traditional practice, the bearings were simply located between the
superstructure (95) and foundation (100) with no bonding at the contact
surfaces. As
a result, the installation of the bearings was significantly easier. Such an
application is
justified where no detachment due to overturning occurs between the
superstructure
and the isolator bearings. Square bearings can be cost effectively produced by
cutting
a long manufactured FREI strip to the required size. Additional cost savings
are
achieved by eliminating the thick end plates which are commonly bonded to the
top
and bottom faces of conventional elastomeric bearings. Also, because of the
relatively
short distance (105) between the adjacent isolators (90) (i.e., 2.40 m as
indicated in
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
Fig. 5b), the concrete beam (115) located at the top of the bearings to carry
the
superstructure between them, is not required to be heavily reinforced.
Moreover, no
special foundation system is required. Accordingly, no significant additional
cost is
imposed on construction of the base isolated building of this embodiment.
[0097] The fundamental natural period of the fixed base model of the masonry
building (110) shown in Fig. 5a, along either directions x or y, was
calculated to be
0.17 s. In this analysis, the in-plane horizontal stiffness of the masonry
shear walls
due to bending and shear, was calculated in a model where each wall was
considered
as a cantilever with both ends fixed against rotation and was subjected to a
horizontal
in-plane load at the top of the storey. To account for cracking in the masonry
shear
walls, an equivalent thickness was evaluated based on the effective moment of
inertia
of ./, = 0.351g as suggested for cracked section properties of beams and walls
by the
ACT code (ACI 318-02, section 10.11.1). Although the equivalent thickness of
the
walls (i.e., te = 0.35t) was taken into account for modelling purposes, the
weight of
the walls was calculated based on their full thickness. The total weight of
the building,
including the base diaphragm, was found to be around 511 kN. Therefore, each
bearing was subjected to approximately 64 kN vertical load.
[0098] The
ratio of the isolated building period to the fixed-base building period is
a measure of the benefit of a base isolation system in hazard mitigation.
Typical
earthquake accelerations are known to have dominant periods of between 0.1 s
and
1.0 s with maximum severity often in the range of 0.2 ¨ 0.6 s. In the
frequency range
of seismic events, the isolated building period is also known as the seismic
isolation
period. Therefore, achieving a target isolated building period larger than 1.0
s, being
more than 5 times longer than the fixed-base period of the building, is deemed
to be
reasonable. In the base isolation system of aspects of the present invention,
the
bearings were subjected to relatively light pressure. Therefore, to achieve
the desired
isolated building period, a soft compound of rubber material can be utilized
in the
construction of these bearings.
21
CA 02727039 2010-12-06
WO 2008/148203 PCT/CA2008/001077
[0099] Fig. 6
illustrates the simple mass-spring-dashpot idealization used to model
the prototype base isolated building. The bearing isolators in the base
isolation system
had nonlinear lateral load-displacement behaviour, along with a 5% equivalent
viscous damping ratio, an experimentally evaluated average value. Since in a
base
isolated system the superstructure remains nearly rigid, it is modeled as
linear elastic.
In order to account for the largest possible response in the superstructure,
this linear
model incorporates the cracked section properties for the superstructure
(i.e., and
te of the shear walls were taken 0.351g and 0.35t, respectively) along with
the 2%
equivalent viscous damping ratio. To understand the dynamic response of the
corresponding fixed-base building, the same linear stiffness properties of the
superstructure along with 5% equivalent viscous damping ratio were taken into
account in the analysis.
[00100] It is understood that performing a time history design analysis will
be
helpful to verify the seismic mitigation effectiveness of the described base
isolation
system. This objective is achieved by comparing the dynamic responses of the
fixed
base and base isolated model of the building subject to a known earthquake
record.
The governing equations for the dynamic system shown in Fig. 6 are as follows:
[1] Ar v.*+ ev*A-K*V* = r* fig
[2] Where,
[ m b [i [M]1k
M* m = C* = rch - K* = [ b 0 õ
[AY} [Al] [0 Kt 0 [K]
{
} * 1i * {vb r} = {1 r ¨
{off v ¨
m = m1 + m2
where, [M], [C] and [K] are mass, damping, and stiffness matrices,
respectively, of
the fixed base model of the superstructure. The vertical component of ground
motion
is ignored in this analysis. Additionally, the influence of variation in
vertical
compression on the lateral load-displacement characteristics of the bearings
is
neglected. The horizontal stiffness of a bearing is generally affected by the
value of
the vertical compression load. However, results have indicated that when the
applied
22
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
vertical compression load is significantly lower than the buckling load of the
bearing,
its influence on the horizontal stiffness of the bearing is negligible.
[00101] The acceleration time history for the NS component of the El Centro
1940
earthquake was selected and scaled by a factor of 1.32 in order to match its
Peak
Ground Acceleration (PGA) to the maximum expected value (with 2% probability
of
exceedence in 50 years) in Vancouver, British Columbia, Canada, namely, 0.46g
(2005, National Building Code of Canada). This simulated accelerogram is
denoted
hereafter as the input earthquake. Comparison between the 5% damped pseudo
acceleration spectrum of this earthquake and the design spectrum for Vancouver
showed that for the periods longer than approximately 0.2 s, the amplification
of this
earthquake was larger than the expected values in Vancouver. This means that
the
response of the base isolated building may become even larger than the
response of
the fixed base model. Values for other locations around the world, including
those
that are known for seismic activity, can be used to test models and bearings
of the
present invention.
[00102] Fig. 7 shows the resulting lateral load-displacement hysteresis of the
designed bearing, with 5% damping ratio for the input earthquake. Due to
sufficient
damping properties of the bearing, the peak of the resulting lateral
displacement is
lower than the lateral displacement capacity of the bearing (i.e., 200% tr 140
mm).
[00103] Fig. 8a shows the time history of the designed bearing's lateral
displacement and Fig. 8b presents the time history of the isolated building
period of
the base isolated system. As indicated, the isolated building period,
depending on the
acceleration amplitude of the input earthquake, fluctuated between 0.96 s and
1.31 S.
This occurred as the result of SR deformation behaviour in the bearings. Such
an
increase in the isolated building period cannot be expected in a FREI bearing
with
bonded boundary conditions at its contact surfaces or in a traditional SREI
bearing. A
SREI even in an unbonded application does not exhibit a SR deformation because
of
the flexural rigidity of the reinforcing layers. These latter bearings perform
almost
linearly. To achieve an isolated building period of 1.31 s, the total height
of FREI-
bearings employed in bonded application or SREI bearings, having the same
rubber
23
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
properties, would need to be increased. An increase in the height of these
bearings is
usually accompanied by an increase in the plan dimensions due to stability
considerations. This implies that FREI bearings in an unbonded application
have
shorter operational height and are significantly more cost effective than
conventional
bonded bearings. Therefore, the base isolation system of the present invention
is an
adopted solution for seismic mitigation of ordinary low rise buildings and
like
structures.
[00104] Fig. 9 presents the absolute accelerations at the 1st floor and the
roof of the
fixed base (FB) and the base isolated (BI) buildings. As indicated, while the
induced
floor acceleration in the FB-model increases with floor height, it remained
nearly
constant in the BI-model. This indicates that the superstructure in the base
isolated
model remained almost undeformed.
[00105] The time histories of base shear as well as the inter-storey drift at
the 2'd
storey of the building in the FB and BI models are presented in Figs. 10a and
10b.
[00106] Table 1 contains the peak response values for the fixed-base building
as
well as the corresponding base isolated building with 5% damped bearings in
the base
isolation system. The significant decrease in the peak response values (around
60%)
of the base isolated system with respect to those of the corresponding fixed
base
model established that the base isolation system of this embodiment performed
efficiently against the considered input earthquake. A minimum safety factor
of 2.3
against overturning moment in the base isolated building was achieved. This
implies
that no detachment due to overturning moment, at the contact surfaces of the
bearings
occurred. Considering the seismic data in Vancouver, BC, a lateral
displacement
capacity of 140 mm in a base isolation system having 5% effective damping is
enough
to achieve an isolated building period of 1.31 s. Accordingly, the proposed
base
isolation system is effective for use in seismic mitigation of buildings and
other
structures to be constructed in these regions, as well as regions of lower
seismic
activity.
24
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
Table 1. Peak response values of the fixed-base
and the base isolated masonry building
Absolute Accel. (g) Inter-storey Drift (mm) Base Shear Overt. Moment
1st floor roof 1st Storey 2nd Storey (kN) (kN.m)
FB model 0.78 1.16 5.7 3.0 334 1462
BI model 0.37 0.38 2.3 1.0 136 556
Response
52.6% 67.2% 59.6% 66.7% 59.3% 62.0%
Reduction
Summary of Example 1
[00107] Test results for this embodiment were conducted on square carbon-FREI
bearings employing soft compound low damped natural rubber as the elastomer.
The
bearings were meant to be used for base isolation in structures such as
ordinary low-
rise residential and commercial buildings or other low-rise structures. The
bearings
were not bonded to the test platens. For bearings having suitable aspect ratio
values
(i.e., NB I and NB2), this particular type of application has resulted in a
stable rollover
(SR) deformation, which reduced the horizontal stiffness and increased the
efficiency
of the bearing as a seismic isolation device. As such these bearings are
denoted as
stable unbonded (SU) FREI-bearings.
[00108] Cyclic tests on SU-FREI Bearings NB1 and NB2 showed that the isolated
period of these bearings, depending on the severity of earthquake, vary
between 1.1 to
1.4 s and 1.0 to 1.2 s, respectively. Furthermore, NB1 and NB2 were tested at
up to
140 mm and 120 mm lateral displacement, respectively. Considering the achieved
base isolated values, these lateral displacement capacities are considered to
be
sufficient for many moderate to high seismic risk regions including along the
west
coast of North America, provided that about 5% and higher effective damping
could
be achieved in the isolation system. Test results establish that application
of the
bearings in one embodiment can significantly reduce the earthquake induced
force
and interstory drift in a masonry shear wall structure with a fixed-base
period in the
range of 0.1 s.
[00109] Vertical testing revealed that carbon fiber reinforcement provides an
acceptable vertical stiffness in FREI bearings. A vertical frequency of around
17 Hz
was achieved in Bearings B1 and B2, which is considered as an acceptable value
for
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
seismic isolation application. Due to the shorter height, Bearings NB1 and NB2
exhibited even larger vertical stiffnesses compared to Bearings B1 and B2.
[00110] Results of a time history analysis performed on a fixed-base as well
as a
base isolated model of a prototype masonry shear wall structure, as a typical
small
low-rise building, are described. In a novel application, unbonded square
carbon FREI
bearings are utilized in the base isolation system of the present invention.
Due to the
relatively close spacing of the bearings, no special structural elements or
foundation
system, and no significant additional cost are required for the construction
of the base
isolated building. Moreover, the unbonded application at these bearings
permits the
use of shorter operational height and eliminates the need for thick steel end
plates. In
this regard, the base isolation system of the present invention is a cost
effective
seismic isolation technique. Analytical results indicated that an unbounded
FREI
bearing with 5% damping can effectively decrease the seismic demand on the
buildings constructed in high seismic risk regions.
Example 2
[00111] Two laminated fiber reinforced elastomeric sheets, denoted as Sheets
B1
and B2, having physical dimensions of 200 x 200 x 25 mm were constructed. Each
sheet consisted of twelve layers of soft unfilled Neoprene (TM) synthetic
rubber
(hardness: 37 5 Durometer, shore A (ASTM D2240, 2005)) that were
interlaminated
with eleven layers of bi-directional (0/90 orientation) carbon fiber fabric as
the
reinforcement. A cold vulcanizing chemical compound (a rubber cement) was
utilized
to bond the elastomer and reinforcement layers together. The matrix volume
fraction,
defined as the relative volume of matrix to the fibers in each fiber
reinforcement
layer, was approximately 54%. Sheets B1 and B2 were subsequently cut to
produce
eight bearings in total. Bearings B1-1 and B1-3, which were originally located
along
the diagonals of Sheet BI, were selected for the prototype testing (see Fig.
11).
Bearings B2-1 and B2-3 were similarly chosen from Sheet B2. These 1/4 scale
model
bearings of the prototype full size bearings had 70 x 70 mm dimensions in
plan. The
total bearing thickness was approximately 25 mm of which the total thickness
of the
twelve 1.58 mm thick rubber layers was tr = 19 mm combined with eleven 0.54 mm
thick reinforcing layers. For a 54% matrix fraction, the implied thickness of
the
26
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
carbon fiber fabric was (1-0.54)(0.54 mm) = 0.25 mm. To prevent any unexpected
premature delamination, two coats of the bonding compound were applied to the
perimeter surfaces of the bearings.
[00112] The supplier-provided nominal tensile modulus of the elastomer at 100%
elongation was 1.2 MPa. This implied a nominal shear modulus of G = 0.4 MPa
for
an incompressible elastomer with a Poisson's ratio of 0.5. Owing to this, the
nominal
effective horizontal stiffness of a bearing with cross section area of A (in
plan), was
estimated by GA/t, to be 103 N/mm. This equation suggests a uniform
distribution of
shear stress. It can be used to approximate the shear stiffness of
conventional SREI
bearings within the practical shear strain range (typically up to 100%), where
experimental observations imply a nearly linear lateral load-displacement
relationship
(1999, Mori A., Moss P. J., Cooke N., and Carr A. J.. "The Behavior of
Bearings
Used for Seismic Isolation under Shear and Axial Load." Earthquake Spectra,
Vol.
15, No. 2, Pages 199-224).
[00113] It is understood that this equation overestimates the value of
effective
horizontal stiffness for a FREI bearing in an unbonded application due to the
additional degree of freedom at the contact surfaces compared to traditional
bearings
in a bonded application (Sept. 2007, Toopchi-Nezhad H., Tait M. J., and
Drysdale R.
G. "Testing and modeling of square carbon fiber reinforced elastomeric seismic
isolators." Journal of Structural Control and Health Monitoring, Published
online).
For a vertical pressure of p = 1.6 MPa, the estimated effective horizontal
stiffness
resulted in an isolated period of 0.55 s for the 1/4 scale bearing. According
to dynamic
similitude law (1999, Harris H. G., and Sabnis G. "Structural Modeling and
Experimental Techniques." CRC Press, New York), the displacement, period, and
stiffness of the full scale model are respectively, 4, 2, and 4 times larger
than the
corresponding values of the 1/4 scale model, respectively. Using the scaling
factor, the
corresponding estimated period is 1.1 s for the full scale bearing. From the
above
estimated period, the initial design of the bearings meet the desired target
isolated
period provided that the bearings exhibited SR-deformation under large lateral
displacements.
27
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[00114] As described above, the aspect ratio (R) and shape factor (S) are two
crucial physical properties of an elastomeric bearing that significantly
influence its
mechanical characteristics. By definition, the shape factor is the ratio of
vertically
loaded area on one face of the elastomer layer to its perimeter load-free
area, and
aspect ratio reflects the ratio of length to total thickness of the bearing.
The shape
factor of the bearings studied was approximately S ¨ 11, and the aspect ratio
was R =
2.8. Aspect ratio plays an important role in achieving SR-deformation in the
bearing.
A smaller aspect ratio makes the FREI bearing more susceptible to rollout
instability.
In this regard, a block-type FRET bearing with an appropriate aspect ratio is
to be
employed in an unbonded application. SR-deformation in a block type FRET
bearing
with unbonded contact surfaces results in a significantly decreased effective
horizontal stiffness. Accordingly sufficient flexibility can be achieved in
the lateral
response of the base isolation system.
[00115] As a result of being square in plan view, the bearings were expected
to
exhibit similar mechanical properties in the two perpendicular directions of
movement. One of the vertical faces of each bearing was marked with the
bearing's
name and deemed as the reference face. The response of the bearing for loading
parallel and perpendicular to the reference face was denoted as 00 and 900
response,
respectively. A 45 response corresponds to the case where the lateral
displacement
was along the diagonal of the bearing. Overall, a bearing's response to
lateral
displacement was categorized as either parallel (i.e., at 00), perpendicular
(i.e., at
900), and diagonal response (i.e., at 45 ).
[00116] All of the bearings were tested under vertical compression (vertical
test).
Additionally, under a prescribed constant level of vertical pressure, the
bearings were
subjected to sinusoidal cycles of lateral displacement (cyclic test) in a
manner like
that described in Example 1, above.
[00117] The time history pattern of lateral displacement in the cyclic testing
portion
of the study was varied in terms of displacement rate (rate) and pattern of
variation in
the amplitude of cyclic lateral displacements. The lateral displacement time
history
pattern had either an increasing amplitude pattern (ascending amp.) or a
decreasing
28
CA 02727039 2010-12-06
WO 2008/148203 PCT/CA2008/001077
amplitude pattern (descending amp.). All tests were conducted at room
temperature.
Table 2 lists the example test sequence and details for each bearing, namely
cyclic
tests with ascending displacement amplitude pattern and rate sensitivity
tests.
Table 2. Test sequence carried out on each bearing isolator*
0
0 Model Scale Bearing
c
4) ______________________________________________________________________
=
Cr
ct B1-1 B1-3 B2-1 B2-3
Cyclic test (00) Vertical test Cyclic test (00) Vertical test
1 ascending amp., p = 0.82, 1.63, 2.45 MPa ascending amp., p
= 0.82, 1.63, 2.45 MPa
rate = 30 mm/s rate = 30 mm/s
Rate sensitivity test Cyclic test (0 ) Rate sensitivity
test Cyclic test (00)
2 0 response, 0 response,
dispL amp. = 100% t, ascending amp., dispi. amp. = 100% t, descending
amp.,
rate = 30, 45, 76, 114 mm/s rate = 76 mm/s rate = 114, 76, 45,
30 mm/s rate = 76 mm/s
Rate sensitivity test Cyclic test (90 ) Cyclic test (90 )
Cyclic test (90 )
3 900 response,
displ. amp. = 150% t, descending amp., ascending amp.,
ascending amp.,
rate = 30, 45, 76, 114 mm/s rate = 76 mm/s rate = 76 mm/s
rate = 76 mm/s
Ink of vert. pressure In& of vert. pressure In& of vert.
pressure In& of vert. pressure
on horz. response on horz. response on horz. response on horz.
response
4 90 response, 90 response, 90 response, 90 response,
p = 1./, /.6, 2.4 MPa, p = 0.8, 1.6, 2.4 MPa, p = 0.8, 1.6, 2.4
MPa, p = 0.8, 1.6, 2.4 MPa,
dispi. amp. = 100% tr, dispL amp. = 100% tr, dispi. amp. = 100%
tr, dispi. amp. = 100% tr,
rate = 30 mm/s rate = 76 mm/s rate = 76 mm/s rate = 76 mm/s
In& of vert. pressure Rate sensitivity test Cyclic test (45 )
Rate sensitivity test
on horz. response
5 90 response, 0 response, 0 response,
p= 1.1, 1.6, 2.4 MPa,
dispi. amp. = 150% t, dispi. amp. = 100% tr, ascending amp.
dispi. amp. = 100% tr,
rate = 30 mm/s rate = 30, 45, 76, 114 mm/s rate = 76 mm/s
rate = 30, 45, 76, 114 mm/s
Cyclic test (45 ) Cyclic test (450) Vertical test Cyclic test (45 )
6 ascending amp., ascending amp., descending amp.,
rate = 30 mm/s rate = 76 mm/s p = 0.82, 1.63, 2.45 MPa rate = 76
mm/s
Vertical test Vertical test Cyclic test (0 ) Vertical test
7 p = 0.82 1.63, 2.45 MPa p = 0.82, 1.63, 2.45 MPa ascending
amp., p = 0.82, 1.63, 2.45 MPa
rate = 76 mm/s
Cyclic test (0 ) Cyclic test (90 ) Cyclic test (0 )
8 ascending amp. ascending amp., ascending amp.,
rate = 76 mm/s rate = 76 mm/s rate = 76 mm/s
Cyclic test (00)
9 ascending amp.,
rate = 76 mm/s
* Unless otherwise stated, cyclic tests were conducted under a constant
vertical pressure of 1.6
MPa.
29
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[00118] The time history pattern of lateral displacement in the cyclic testing
was
varied in terms of average displacement rate (rate) and pattern of variation
in the
amplitude of cyclic lateral displacements. The lateral displacement time
history
pattern had either an increasing amplitude pattern (ascending amp.) or a
decreasing
amplitude pattern (descending amp.). Lateral cyclic tests were conducted under
displacement control with vertical load controlled to maintain a constant
vertical
compressive force. Each bearing was subjected to fully reversed sinusoidal
cycles of
lateral displacement having amplitudes of 25% tõ 50% tõ 75% tõ 100% tõ 150%
tr,
and 200% tr. Considering the total thickness of rubber layers in each bearing
(t, = 19
mm), these correspond to lateral displacement amplitudes of 4.8, 9.5, 14.3,
19.0, 28.5,
and 38.0 mm, respectively.
[00119] When displaced under pure shearing action, a FREI bearing in an
unbonded
application, exhibits rollover deformation. One study (Toopchi-Nezhad et. al.
2007))
revealed that for a FREI bearing employed in an unbonded application, given
the
material properties and shape factor (S), the aspect ratio (R) plays an
important role in
achieving SR-deformation in the bearing. No closed form prior art solution
exists to
evaluate the proper aspect ratio that leads to SR-deformation, based on the
physical
geometry and material properties of the FREI bearing. However, results from
lateral
cyclic tests on prototype bearings having identical characteristics but
different aspect
ratios, can be used to determine a suitable aspect ratio in the final design
of the
bearing.
[00120] As a primary study only 1/4 scale Bearing B1-1, with an aspect ratio
of R =
2.8, was cut from Sheet B1 (see Fig. 11). The objective was to verify the
suitability of
the selected aspect ratio in achieving SR-deformation in the bearing.
Furthermore,
there was a need to verify the adequacy of the resulting effective horizontal
stiffness
and damping values at different amplitudes of lateral displacement, through
examining the cyclic test results.
[00121] Bearing B1-1, at its virgin state, was initially subjected to a set of
cyclic
tests, at 00 orientation, with different amplitudes of lateral displacement.
For each
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
amplitude of displacement, the bearing was loaded to p = 1.6 MPa vertical
pressure. It
was subjected to four fully reversed sinusoidal displacements with identical
amplitudes and then vertically unloaded. The average rate of cyclic
displacements was
approximately 30 mm/s. This bearing remained stable and showed acceptable
behavior over the entire test range of lateral displacement amplitudes. Next,
Bearing
B2-1 with the same physical dimensions as B1-1 was cut from Sheet B2 and was
subjected to cyclic lateral displacements at 00 orientation under constant 1.6
MPa
vertical pressure. As before, the selected displacement rate was 30 mm/s.
However,
unlike the case of Bearing B1-1, only 3 cycles of lateral displacement were
performed
at the 6 displacement amplitudes ranging from 25% to 200% tr and the constant
vertical pressure was maintained on the bearing without unloading between the
various displacement amplitudes. As an example, only the hysteresis loops for
Bearing B2-1 (at 00), corresponding to the final cycle of lateral
displacements at each
displacement amplitude, are shown in Fig. 12a where, similar to B1-1,
acceptable
performance was observed.
[00122] The diagonal response of Bearing B1-1 was also investigated through a
test
sequence similar to the 00 orientation test for Bearing B1-1. Fig. 12b shows
the
resulting hysteresis loops corresponding to the final cycle for each
displacement
amplitude. The diagonal cyclic test usually resulted in the most distorted
pattern of
deformation in the bearing. After completion of the diagonal test, partial
debonding in
one of the laminates near midheight of Bearing B1-1 was observed. The
delamination
existed over a length of approximately 1 mm on one face of the bearing and was
difficult to see. As shown in Fig. 12b, the bearing remained stable and
demonstrated
an acceptable lateral response even during the last cycle of displacement.
[00123] From Figs. 12a and 12b, it was observed that, due to rollover
deformation
in the bearing, the effective horizontal stiffness decreased with increased
amplitude of
lateral displacement. However, positive values of the tangent horizontal
stiffness
throughout the hysteresis loops implied that the bearings showed an acceptable
SR-
deformation. Table 3 shows the test results, discussed above, for Bearings B1-
1 (at 00
and 45 ) and B2-1 (at 0 ).
31
CA 02727039 2010-12-06
WO 2008/148203 PCT/CA2008/001077
Table 3. Effective horizontal stiffness values and damping ratios
corresponding to
various lateral displacement amplitudes in Bearings B1-1 and B2-1*
B1-1, 00 (virgin) B2-1, 00 (virgin) B1-1, 450
Displacement __________________________________
Stiffness Damping Stiffness Stiffness Stiffness Damping
Amplitude
(N/mm) (%) (N/mm) (N/mm) (N/mm) (%)
st _____________________________________________
115.5 12.5 135.7 12.7 105.9 13.2
cycle
25% tr
Last
** **
112.1 11.1 130.4 11.1 105.5 12.3
cycle
1st
96.8 10.4 107.4 9.7 83.4 10.8
cycle
50%t1
Last
91.6 8.7 102.6 8.5 82.4 9.8
cycle
1st
82.3 9.3 90.7 8.5 71.2 10
cycle
75% tr
Last
77.2 7.9 86.4 7.5 70.3 8.8
cycle
1s`
72.5 8.7 79.2 8.0 64.1 9.8
100% cycle
tr Last
68.3 7.4 76.0 7.1 63.1 8.5
cycle
lst _____________________________________________
59.2 9.7 65.2 8.7 54.6 10.9
150% cycle
tr Last
53.8 7.6 60.4 6.9 52.9 9
cycle
1st
54.2 9.5 61.0 8.5 47.7 12.9
200% cycle
tr Last
47.0 8.6 54.2 7.1 43.9 11.4
cycle
*Average rate of lateral displacements for all bearings cited in this Table =
30 mm/s
32
CA 02727039 2014-09-26
These values belong to 27.5% tr lateral displacement amplitude as this was the
initial test. The subsequent bearings were tested at 25% tr.
[00124] In general, the lateral response of Bearings B1-1 and B2-1 appeared to
be
consistent. Considering the resulting effective horizontal stiffness values,
the
corresponding isolated periods of the tested 1/4 scale bearings range between
0.5 to 0.8
s depending on the lateral displacement amplitude. Therefore, the bearing
isolated
periods of the corresponding full scale bearing are expected to range from 1.0
to 1.6 s
which is either equal to or beneficially longer than the target isolated
period.
[00125] At 100% elongation, the rubber supplier specified a nominal damping of
the elastomer of approximately 5%, whereas significantly larger damping values
ranging from 7% to 13% in both the parallel and diagonal response were
achieved,
depending to the displacement amplitude. This implies that the fiber
reinforcement is
a new source of energy dissipation in a FREI bearing. Bearing B1-1 performed
more
effectively in the diagonal response (45 ) than the parallel response (0 ) due
to the
lower effective horizontal stiffness and larger damping values achieved.
Consequently, the selected aspect ratio of R = 2.8 was deemed to be adequate
and the
preliminary design of the bearings was verified.
[00126] For rate sensitive isolators, cyclic testing with sinusoidal lateral
displacements should be conducted with an excitation frequency equal to the
natural
frequency of the isolator at the design displacement (2005, American Society
of Civil
Engineers. "Minimum design loads for buildings and other structures." ASCE/SEI
7-
05). According to the cyclic test results on Bearings B1-1 and B2-1 (see Table
3), at
the design lateral displacement (i.e., around 100% tr = 19 nun), an isolated
period of
approximately 0.67 s will be achieved. This implies a natural frequency of 1.5
Hz and
an average displacement rate of 114 minis in the input sinusoidal cycles of
lateral
displacement. Although the lateral actuator used in the test setup was able to
operate
at this rate at 19 mm ( 100% tr) lateral displacement, it was not able to
operate at
this rate at higher amplitudes. Therefore, a lower rate that would still
provide results
sufficiently accurate from an engineering perspective was required. This
section
discusses the procedure followed, leading to the selection of a suitable
displacement
rate.
33
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[00127] All of the model bearings, under constant 1.6 MPa vertical pressure,
were
subjected to three fully reversed sinusoidal cycles of 100% tr = 19mm lateral
displacement sequentially at average rates of 30, 45, 76, and 114 mm/s. A 1.6
MPa
constant vertical pressure was maintained. As mentioned above, the average
rate of
114 mm/s was associated with the natural frequency of the bearing at the
design
displacement and was chosen as the reference rate for comparison. The
influence of
displacement rate on effective horizontal stiffness and damping ratio of the
bearings
can be seen in Figs. 13a and 13b, respectively.
[00128] Overall, both effective stiffness and damping increased with increased
displacement rate. Bearing B2-1 showed the most sensitivity to the
displacement rate
with approximately a 9% reduction in the effective stiffness and more than 16%
reduction in the damping of the bearing at the lowest rate (30 mm/s). Unlike
Bearing
B2-1, cyclic tests for all other bearings started with the lowest rate and
ended at the
highest rate. As a result, at displacement rates lower than 114 mm/s, these
bearings
showed less than 6% and 11% reduction in the effective stiffness and damping,
respectively. A rate sensitivity test on Bearing B1-1 was also carried out at
150% tr
(28.5 mm) lateral displacement. Test results showed a similar trend of
reduction in
stiffness and damping at lower displacement rates.
[00129] In order to work within the range of equipment capability, an average
displacement rate of 76 mm/s was selected. This rate resembled an excitation
frequency of approximately 1 Hz at 100% tr lateral displacement which was
sufficiently close to the bearings' natural frequency (i.e., approximately 1.5
Hz). At
this rate, as the worst case scenario (results of Bearing B2-1) as shown in
Fig. 13a
and 13b, respectivelyõ the effective horizontal stiffness and damping at each
cycle
are on average 6% and 10% lower than the corresponding values at the natural
frequency of the bearings. These variations are considered minor and can be
neglected
for many practical applications.
[00130] A time history plot of the input excitation, comprising six
displacement
amplitudes in an ascending pattern, with a displacement rate of 76 mm/s is
shown in
Fig. 14a. The displacement history contains 3 fully reversed cycles at
amplitudes of
34
CA 02727039 2010-12-06
WO 2008/148203 PCT/CA2008/001077
25%, 50%, 75%, 100%, 150%, and 200% tr (where t, = 19 mm). Bearing B1-3, at
its
virgin state, was initially subjected to this displacement history, which was
imposed at
0 orientation on the bearing. Fig. 14b shows the resulting hysteresis loops
for all
cycles of the test and Table 4 contains the corresponding properties of this
bearing.
Table 4. Parallel and perpendicular responses of Bearings B1-3, B2-1, and B2-
3*
B1-3, 0 (virgin) B2-1, 90 B2-3, 900
Displacement
Cycle Stiffness Damping Stiffness Damping Stiffness Damping
amplitude
(N/mm) (%) (N/mm) (%) (N/mm)
(c/o)
1 164.0 13.7 142.7 12.4 136.5 12.5
25%t, 2 155.6 13.4 139.8 11.8 134.0 12.4
3 154.1 13.3 138.8 12 133.2 12.8
1 123.3 11.9 109.1 11.2 105.3 11.2
50% tr 2 116.5 10.9 104.8 10.5 101.9 10.5
3 114.6 10.8 103.8 10.4 100.8 10.7
1 103.3 10.2 89.9 9.8 86.3 9.8
75% t, 2 97.1 9.5 86.2 9.5 83.3 9.6
3 95.3 9.4 85.2 9.4 82.4 9.5
1 89.3 9.4 78.6 9.1 75.0 9.1
100% tr 2 84.7 8.8 75.8 8.8 73.0 8.9
3 83.6 8.7 75.2 8.7 72.3 8.7
1 72.9 9.9 65.7 9.2 61.8 9.3
150% tr 2 68.8 8.4 63.2 8.3 59.8 8.4
3 67.1 8.3 62.2 8.1 58.9 8.2
1 68.5 9.4 65.6 8.7 61.2 8.7
200%t1 2 63.3 8.3 61.0 7.7 57.4 8.0
3 61.6 8.1 60.0 7.4 56.4 7.8
*Average rate of lateral displacements for all bearings cited in this Table =
76 mm/s
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[00131] As seen in Fig. 14b, as with Bearings B1-1 and B2-1, the hysteresis
loops
showed acceptable response characteristics. However, the resulting mechanical
properties are more accurate as they incorporate a more representative
displacement
rate in the cyclic testing. Figs. 15a to 15d contain photographs of Bearing B1-
3 at
various displacement amplitudes for cyclic testing in the parallel (0 )
direction.
Bearings B2-1 and B2-3 were subjected to the same displacement history
presented in
Fig. 14a along their perpendicular (90 O) orientation. Table 4 contains the
resulting
mechanical characteristics of these bearings.
[00132] During the first cycle of lateral displacements with constant
amplitude, all
bearings tested in this study exhibited higher stiffness and damping than in
the
following cycles (see Fig. 14b and Table 4). When an elastomer in its virgin
condition
is subjected to any constant amplitude of cyclic tension, compression or shear
loading,
reloading of the elastomer up to the maximum previously achieved displacement
amplitude usually requires lower stress than the initial cycle. The maximum
difference in the required reloading stresses exists between the first and the
second
cycles and becomes negligible after 6-10 cycles depending on the elastomer
compound (2004, Dorfmann A., Ogden R.W. "A constitutive model for the Mullins
effect with permanent set in particle-reinforced rubber." International
Journal of
Solids and Structures, 41, Pages 1855-1878)). This phenomenon, which is
observed
in both unfilled and particle-reinforced rubber, is known as the "Mullins
effect"
(2002, Marckmann G., Verron E., Gornet L., Chagnon G., Charrier P., and Fort
P. "A
theory of network alteration for the Mullins effect." J. Mech. Phys. Solids
50, Pages
2011-2028). In all bearings tested, at each lateral displacement amplitude,
after the
first cycle, the hysteresis loops stabilized so that stable effective
horizontal stiffness
and damping values in the 2nd and ri cycles were achieved. Such stable
properties are
usually denoted as "scragged" properties. The untangling or breakage of weak
crosslinks in an unfilled rubber can be considered as a micro-mechanical
interpretation of this phenomenon. Depending on the elastomer compound, it is
known that partial recovery of unscragged properties with time is likely to
occur.
[00133] Figs. 16a and 16b present the variations of lateral response with the
amplitudes of cyclic lateral displacement for the bearings listed in Table 5.
The values
36
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
in Fig. 16 represent the average response of all cycles at each displacement
amplitude
of cyclic testing. As indicated in Fig. 16, due to rollover deformation, the
effective
horizontal stiffness significantly decreased with increased amplitude of
lateral
displacement. However, the reduction in damping at displacements larger than
100%
tr was approximately 10% which is considered to be negligible. This indicates
that the
bearings would effectively reduce the shear force transmitted to the
superstructure,
while maintaining a sufficient amount of damping to limit the bearing's
lateral
displacement against severe excitations.
Table 5. Diagonal response of Bearings B1-3 and B2-1*
B1-3, 450 __________________________________________ B2-1, 450
Displacement
Cycle Stiffness Damping Stiffness Damping
Amplitude
(N/mm) (%) (N/mm) (%)
1 135.7 15.8 130.2 14.0
25% tr 2 131.5 16.1 130.5 14.2
3 130.6 16.1 129.7 14.2
1 102.5 12.1 101.9 11.2
50%tr 2 101.0 11.5 99.8 10.3
3 99.9 11.4 99.0 10.1
1 84.0 10.7 83.7 9.8
75% tr 2 82.3 10.4 81.7 9.4
3 81.9 10.2 81.1 9.4
1 72.3 10.2 72.9 9.4
100% tr 2 71.3 9.7 71.6 8.8
3 70.8 9.7 71.2 8.8
1 58.7 10.9 59.8 10.0
150% tr 2 57.7 10.0 59.1 8.9
3 57.1 9.8 58.3 8.9
1 51.9 11.7 54.5 10.6
200% tr 2 49.4 11.0 51.8 9.4
3 48.4 10.9 50.5 9.2
*Average rate of lateral displacements for all bearings cited in this Table =
76 mm/s
37
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[00134] Unlike Bearing B1-3, Bearings B2-1 and B2-3 had been subjected to a
previous set of cyclic tests in the parallel (0 ) direction (see Table 2).
Accordingly, as
indicated in Table 4, while damping values are comparable, Bearings B2-1 and
B2-3
showed lower effective horizontal stiffness than B1-3.. Bearing B2-3 showed
more
response reduction in the perpendicular direction (90 ) than Bearing B2-1.
This
bearing had been subjected to previous cyclic testing with a descending
pattern of
displacement amplitudes starting at 200% tr and ending at 25% tr. As a result,
such a
lateral displacement history may have resulted in more damage to the elastomer
crosslinks.
[00135] Bearings B1-3 and B2-1 were also subjected to diagonal cyclic testing,
where the displacement history was applied along the diagonal of the bearings
(45
orientation). The test results are listed in Table 5 and Fig. 17 contains
photographs of
Bearing B1-3 at various displacement amplitudes.
[00136] Similar to the parallel and perpendicular responses (0 or 90 ) of the
bearings, the response characteristics at 45 orientation showed SR-
deformation and a
similar pattern of variation with the displacement amplitude (See Fig. 18).
However,
being scragged, stable hysteresis loops at each displacement level were
achieved. The
elastomer showed minor unscragged properties at the first cycle of 200% tr
displacement. After completion of the test program no visible damage was
observed
in Bearings B1-3, B2-1, and B2-3.
[00137] In this section, the adequacy of both the parallel/perpendicular and
diagonal
responses of the tested bearings against both 0.4g and 0.5g levels of 5%
damped
spectral accelerations at 1 s period (SA,n) were verified. For the purpose of
this testing,
it was presumed that the site class effects (soil properties) are accounted
for in these
spectral values.
[00138] The equivalent lateral-force procedure in ASCE 7 (2005, American
Society
of Civil Engineers. "Minimum design loads for buildings and other structures."
ASCE/SEI 7-05) can be used for the design of many seismically isolated low-
rise
38
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
buildings with regular configuration. This approach takes advantage of the
fact that, in
this particular type of base isolated building, the first mode of vibration
involves
lateral displacements almost entirely at the isolation level, while the
superstructure
remains nearly rigid. Owing to this, the verification procedure can be started
by
evaluating the design lateral displacement DD and corresponding effective
period TD
given by the following expressions.
D gSDITD (3)
D 471,2 BD
TD ¨ 27r _______
kDining
W (4)
where, S Di = 2/3 S ml, BD accounts for the effective damping of the isolator
at DD, W
represents the effective seismic weight (vertical load) on the isolator, and
kpflun is the
minimum effective horizontal stiffness of the isolator at DD, obtained from
cyclic
testing. At MCE, Dm and TM are calculated based on a similar set of
expressions,
however, the subscript D is replaced with M so the variables involved
represent the
corresponding values at MCE.
[00139] Given SDI, an iterative procedure using Eqs (3) and (4) was utilized
in order
to evaluate the DD and TD values of the base isolated system. First, an
initial estimate
of DD was made. Next, the equivalent full scale value of kDmin was obtained
from
Tables 3 or 4 through linear interpolation between the stiffness values
corresponding
to the lateral displacement levels that bracket the current value of DD.
Additionally,
for DD > 200% tr, the test results at 200% tr are utilized. Substituting W and
kanfn in
Eq. (4), TD is evaluated and DD is updated using Equation (3). Iterations are
carried
out until DD and TD converge to their unique values. It is noted that the
vertical
pressure of p = 1.6 MPa implies W = 125.4 kN for the full scale bearing.
[00140] Parallel/Perpendicular Response Evaluation: Conservatively, the
effective
damping in the tested bearings at DD and Dm were assumed to be 8% and 7%,
respectively (see Table 4). This yields values of BD = 1.12 for DBE and BM =
1.08
39
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
corresponding to MCE (ASCE 7, 2005). Additionally, to ensure a conservative
design, the effective horizontal stiffness values, corresponding to the third
cycle of the
test, for Bearing 82-3 tested at 900 were utilized (see Table 4). At a
specific site with
Sm./ = 0.4 and SDI = 0.27, the iterative procedure described above at DBE
leads to a
base isolated period of TD = 1.33 sand lateral displacement of DD= 80 mm (;---
105%
tr) at the center of rigidity of the base isolation system. At MCE this
results in a base
isolated period of TM= 1.48 sand a lateral displacement of Dm = 136 mm (z,'
180% tr).
Similar calculations carried out for a region with Smi = 0.5 and SD! = 0.33
gave values
of TD= 1.43 sand DD= 106 mm (=-' 140% tr) at DBE and TM= 1.50 sand Dm = 172
mm (,---- 226% t1) at MCE.
[00141] Diagonal Response Evaluation: According to Table 5, for lateral
displacements larger than 100% tõ the damping in Bearings B1-3 and B2-1 ranges
from 9% to 11%. Therefore, a conservative value of 9% effective damping,
against
both the DBE and MCE, was assumed. This lead to BD= BM = 1.16 (ASCE 7, 2005).
A similar iterative process, which accounts for the minimum effective
horizontal
stiffness values (interpolated from Table 5), was carried out. For a spectral
acceleration of SDI = 0.27 at DBE, TD= 1.34 s and DD= 76 mm (100% tr) and at
MCE
with SA,11 = 0.4, TM= 1.55 sand DM= 133 mm (175% tr). Similarly, at a specific
site
with Sim/ = 0.5 and SDI = 0.33, at DBE, TD= 1.44 s and DD= 103 mm ("z 136% tr)
and
at MCE, TM= 1.61 s and Dm= 173 mm (c-z; 228% tr).
[00142] A close examination revealed that at 200% tr lateral displacement, the
originally vertical faces of the bearings had not fully contacted the upper
and lower
platens. Therefore, this lateral displacement was not the full contact lateral
displacement (SO of the bearings. Results from cyclic tests conducted on a
bearing
with similar properties to the bearings tested in this research program showed
that the
bearing was able to safely accommodate lateral displacements greater than 250%
tr.
Consequently, the bearings are able to perform satisfactorily under the levels
of
seismic hazard specified herein.
[00143] Significant seismic mitigation is expected in low-rise buildings
(having
fixed base periods of around 0.1 to 0.2 s), that are seismically isolated
using the
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
proposed bearings. If, due to enormous torsional effects or higher earthquake
hazards,
a larger lateral displacement capacity in the bearings is needed, the
bearing's design
can be modified accordingly, based on the description provided herein. Results
of
vertical testing showed that vertical frequency of the 1/4 scale bearings
exceeded 40
Hz. This is equivalent to 20 Hz for the full scale bearings, which is more
than
sufficient as it is significantly higher than the maximum resulting base
isolated
frequency (i.e., 1/T,õ,,, = 1/(1.33 s) = 0.75 Hz). As a result, the
contribution of the
rocking modes in the dynamic response of the base isolated structure becomes
insignificant.
Summary of Example 2
[00144] Lateral cyclic testing was conducted on 1/4 scale carbon Fiber
Reinforced
Elastomeric Isolator (FRET) bearings. The bearings were employed in an
unbonded
application as they were not attached to the upper and lower platens of the
test
machine. Since the fiber reinforcement has no appreciable flexural rigidity,
the
unbonded application resulted in a SR-deformation in the bearings when they
were
laterally deformed. The full contact vertical face lateral displacement (of)
of the
bearings occurred when their originally vertical faces completely came in
contact the
upper and lower platens. Rollout instability lateral displacement, which is
denoted as
5max, is expected to be significantly larger than 8 fc . All tested bearings
exhibited a
positive incremental load-resisting capacity throughout their lateral load-
deflection
response. The bearings showed damping ratios larger than the supplier-
specified
inherent damping of the utilized elastomer. The effective horizontal stiffness
and
damping increased with an increase in the rate of lateral displacements.
However, the
effective horizontal stiffness was found to be less sensitive than the damping
to
variations in the lateral displacement rate.
[00145] Although a soft unfilled (low-damped) Neoprene (TM) was used as the
elastomer, the lateral response of the tested FRET bearings in an unbonded
application
was found to be, in general, similar to that of conventional high-damped steel
reinforced elastomeric bearings. The response can be characterized by larger
effective
horizontal stiffness and damping values at lower lateral displacement levels,
which
41
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
tend to minimize response under wind load and minor seismic excitations. Over
the
lateral displacement range of 100% to 200% tr, the effective horizontal
stiffness is
significantly decreased due to rollover deformation in the bearing. The
decreased
effective horizontal stiffness increases the isolated period of the bearing,
which
improves its efficiency as a seismic isolator. At larger lateral displacements
(larger
than 200% tr) when the originally vertical faces of bearing fully contact the
upper and
lower platens, a significant stiffening in the hysteresis loops is observed.
This
effectively places a limit on the lateral displacements that can occur under
unanticipated seismic excitation levels. As such, efficient seismic mitigation
was
achieved with the tested bearings.
[00146] For those bearings that were subjected to identical loading history,
the
variability of the observed response was relatively low. This can be observed
by
comparing test results for Bearings B1-1 and B2-1 in Table 3, Bearings B2-1
and B2-
3 in Table 4, and bearings cited in Table 5. The predicted nominal effective
horizontal
stiffness of the bearings (i.e.,GA/t, = 103 N/mm) was found to be in close
agreement
with experimental results at lateral displacements of approximately 50% tr.
Accordingly, in the preliminary design, it can be used as a conservative
approximation for the horizontal stiffness of the bearings at design lateral
displacements (i.e., 100% to 150% tr).
[00147] The exploitation of square FREI bearings in an unbonded application is
possible based on this invention. According to the design provisions in ASCE 7
(2005), the tested bearings may be effectively used in seismic mitigation of
many
low-rise buildings with fixed base period of around 0.2 s that are constructed
in
seismic regions with a 5% damped spectral accelerations at 1 s period (Smi) of
up to
0.5g. The design of the FREI bearings may be modified to permit their use in
regions
of higher seismic acceleration. Unique advantages, such as superior damping
properties, the possibility for mass production, simplicity of cutting, ease
of unbonded
installation, and elimination of the need for the thick steel end plates
highlights the
significant potential of this base isolation system to be applied as a cost
effective
method to mitigate the seismic response of many ordinary low-rise buildings
worldwide.
42
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
Example 3
[00148] The carbon SU-FREI bearings of Example 3 were identical to those of
Example 2. The influence of a number of different parameters on the lateral
additional
dynamical variables were studied in this example.
Influence of lateral displacement amplitude history
[00149] As indicated in Table 2, Bearings B1-3, B2-1 and B2-3 were subjected
to
cyclic sinusoidal lateral displacements with an Ascending Amplitude (AA)
pattern
ranging from 25% tr to 200% tr (see Fig. 14a). Bearings B2-3 and B1-3 were
tested
with a Descending Amplitude (DA) pattern of lateral displacements starting at
200%
tr and ending at 25% tr (see Fig. 14c). Consistently, the 1.6 MPa constant
vertical
pressure was applied to the bearings throughout the cyclic testing.
Additionally, the
average rate of lateral displacements for all cases was 76 mm/s.
[00150] By definition, the term "virgin" in this paper applies to the bearings
which
had not been previously tested in any lateral direction. Figs. 14(b) and 14d
show the
resulting lateral load-displacement hysteresis loops at 00 for the virgin
Bearings B1-3
and B2-3. As seen in Fig. 14b, the first cycle of lateral displacements with
constant
amplitude was associated with unscragged response with larger load-resisting
capacity and damping than the following cycles. Once scragged at each
displacement
amplitude, the bearing exhibited stable loops. Additionally, at each lateral
displacement amplitude, significant stiffening following the first cycle was
observed.
However, similar phenomena were not observed in the bearing's response to a
lateral
displacement history with a DA-pattern. As shown in Fig. 14d, the initial
cycle with
the displacement amplitude of 200% tr, scragged the virgin Bearing B2-3 so
that
stable hysteresis loops were achieved for all subsequent cycles with the same
or
smaller displacement amplitudes. Additionally, the hysteresis loops did not
show any
stiffening at the extreme ends of displacement amplitudes lower than 200% tõ
which
is observed when comparing Figs. 14b and 14d.
43
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[00151] In general, compared to a DA-pattern of lateral displacements, an AA-
pattern resulted in larger effective horizontal stiffness. However, at 200% tr
which
corresponds to the largest amplitude, the response properties were found to be
very
similar and thus independent of the history of lateral displacement.
[00152] Regardless of the lateral displacement history and direction of
lateral
displacement, all of the bearings showed a positive incremental load-resisting
capacity throughout the lateral response. Additionally, at the largest lateral
displacement amplitude, when the originally vertical faces of the bearings
(during the
parallel or perpendicular response) contacted the platens of the test machine,
significant stiffening was observed. Therefore, the resulting rollover
deformation was
considered as an acceptable SR-deformation.
[00153] Figs. 19a and 19b contains the lateral response properties of the
bearings
discussed above. In general at the 45 orientation, the bearings exhibited a
lower
effective stiffness in conjunction with a higher effective damping compare to
the
parallel or perpendicular responses. This suggests that the seismic efficiency
of the
bearings at a 45 orientation (diagonal direction) may even be superior to
the parallel
or perpendicular direction. Overall, lateral response to a DA-pattern of
displacements,
whether at 00, 900 or 45 0, was accompanied with a lower effective stiffness
and a
higher damping compare to the corresponding response values for AA-pattern of
lateral displacements.
[00154] The virgin Bearing B1-3 loaded at 0 , provided the upper bound to
the
effective horizontal stiffness, regardless the orientation of lateral
displacements. In the
cyclic test with a DA-pattern, whereas Bearing B2-3 was virgin, Bearing B1-3
had
been subjected to previously cyclic testing in its transverse direction (see
Table 2).
Nevertheless, an excellent consistency between the responses of both bearings
was
achieved. The maximum discrepancy between the average effective horizontal
stiffness values of the two bearings was approximately 3%, which occurred at
200% t-
lateral displacement. As seen in Fig. 23b, Bearing B2-3 (at 450 loading)
provided
the lower bound of the effective horizontal stiffness. Except at the largest
amplitude,
where the bearings were not completely scragged, this bearing showed good
44
CA 02727039 2010-12-06
WO 2008/148203 PCT/CA2008/001077
consistency with Bearings B2-3 (at 00 loading) and B1-3 (at 900 loading). In
the
cyclic test with a DA-pattern of lateral displacements, the resulting
effective
horizontal stiffnesses were consistent (as the initial cycle had scragged the
bearings).
As such, so long as no significant damage occurs in the bearings, any
discrepancy in
the bearings' test results under identical lateral displacement history and
test
procedure is primarily related to the variation in the unscragged properties
of the
elastomer. Accordingly, due to the identical material properties, physical
geometry,
and manufacturing process of the bearings, test observations for the
individual
bearings (including the virgin bearings) under different lateral displacement
histories
can be largely extended to all other bearings.
[00155] Table 7 contains the upper and lower lateral response bounds of the
bearings at 00 or 900 and 450 loading orientations. For the 1.6 MPa vertical
design
pressure and the values of effective horizontal stiffness given in Table 7,
the base
isolated period of the 1/4 scale bearings for lateral displacements ranging
from 25% to
200% tõ depending to the energy magnitude of the earthquake, would range
between
0.4 to 0.7 s for the parallel/perpendicular direction. In the diagonal loading
direction,
the isolated periods would range between 0.5 to 0.8 s. The higher the
earthquake
energy, the longer would be the base isolated period. It is noted that the
corresponding
full scale isolated periods would be about double these values.
Table 7. Lateral response bounds at parallel/perpendicular and
diagonal directions for the tested bearings
Displacement Orientation
Parallel/Perpendicular Diagonal
Displacement
=
Amplitude Cycle Virgin B1-3, 00 B1-3, 900 B1-3, 450
B2-3, 450
(AA - Pattern) (DA - Pattern) (AA - Pattern)
(DA - Pattern)
K (N/mm) (%) K (N/mm) (%) K (N/mm) (%) K (N/mm) (%)
1 164.0 13.7 108.5 11.2 135.7
15.8 97.9 11.0
25% tr
(4.75 mm) 2 155.6 13.4 108.0 9.8 131.5
16.1 101.1 11.5
3 154.1
13.3 108.4 10.1 130.6 16.1 101.9 10.9
123.3 11.9 86.5 11.8 102.5 12.1 81.7 11.7
50% tr
(9.50 mm) 2 116.5 10.9 85.9 11.4 101.0
11.5 81.8 11.4
3 114.6 10.8 86.2 11.1 99.9 11.4
82.0 11.3
103.3 10.2 72.1 10.7 84.0 10.7 68.7 10.7
750 tr 2 97.1 9.5 72.2 10.7 82.3 10.4 68.7 10.5
(14.25 mm)
3 95.3 9.4 72.4 10.7 81.9 10.2 68.8
10.6
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
89.3 9.4 62.5 10.6 72.3 10.2 59.8 10.2
100%tr
(19.00 mm) 2 84.7 8.8 62.8 10.1 71.3 9.7 60.6
10.1
3 83.6 8.7 63.0 10.0 70.8 9.7 60.7 10.1
1 0t 72.9 9.9 50.8 10.4 58.7 10.9 49.1 9.9
50r
(28.50 mm) 2 68.8 8.4 50.6 10.1 57.7 10.0
49.1 9.8
3 67.1 8.3 50.9 10.1 57.1 9.8 49.4 9.6
1 68.5 9.4 64.7 10.8 51.9 11.7 52.5 12.4
200%tr
(38.00 mm) 2 63.3 8.3 61.4 8.6 49.4 11.0 50.1
10.0
3 61.6 8.1 60.2 8.4 48.4 10.9 48.8 9.5
AA - Pattern: Lateral displacements with an ascending amplitude pattern.
DA - Pattern: Lateral displacements with a descending amplitude pattern.
Influence of lateral displacement rate
[00156] Cyclic lateral testing with an average displacement rate of 30 mm/s
was
conducted on Bearings B1-1 and B2-1 (at 0 loading). Considering the 1.6 MPa
vertical design pressure, the natural frequency of the bearings at 100% t, (19
mm)
lateral displacement was found to be approximately 1.5 Hz from the cyclic test
results. This provides an average rate of 114 mm/s for the sinusoidal
displacements
with 19 mm amplitude in the cyclic lateral testing. This rate was denoted as
the
"reference rate". As indicated in Table 2, rate sensitivity tests were
conducted on the
SU-FREI bearings to study the influence of displacement rate on their lateral
response. At 1.6 MPa vertical pressure, three cycles of lateral displacements
having
amplitude of 100% t, were applied to all of the bearings. The lateral
displacements
were applied at different average rates of 30, 45, 76, and 114 mm/s.
[00157] Table 6 contains results of the rate sensitivity tests. It was found
that both
the effective horizontal stiffnesses and damping ratios of the bearings
increased with
the increased displacement rate. However, horizontal stiffness was not as
sensitive as
damping to the displacement rate. Unlike all other bearings, rate sensitivity
tests for
Bearing B2-1 started at the highest displacement rate of 114 mm/s and ended at
the
lowest rate of 30 mm/s. Accordingly, its response appeared to be more
sensitive to
displacement rate. At average displacement rate of 30 mm/s, a maximum
reduction of
10% and 16% in the stiffness and damping of the bearings, respectively, was
observed
compared to the corresponding values at the reference rate. To work within the
test
equipment's range of capability, an average displacement rate of 76 mm/s was
selected for all of the amplitudes of the cyclic test. At this rate, the
influence of rate
46
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
on the lateral response can be neglected as the effective stiffness and
damping were in
average 6% and 10% lower than the corresponding values at the reference rate
of 114
mm/s (Bearing B2-1 in Table 6).
Table 6. Results of rate sensitivity tests at 100% tr lateral displacements
Amplitude of lateral Average rate of sinusoidal lateral displacements
(mm/s)
displacements =
100% tr (19 mm) 30 45 76 114
Bearing
(kN/mm) (%) (kN/mm) CYO (kN/mm) (%) (kN/mm) ( /0)
B1-1 1st Cycle 60.9 10.2 60.6 9.7 61.2 9.9 61.5
10.1
Last Cycle 58.6 8.6 59.4 8.7 59.9 9.0 61.2
9.0
B1-3 1st Cycle 66.7 10.1 66.7 10.1 68.4 10.3
69.8 11.2
Last Cycle 64.2 9.3 65.5 9.4 65.9 9.6 68.3
10.3
B2-1 1st Cycle 66.0 8.5 67.5 8.8 68.5 9.1 74.2
10.6
Last Cycle 65.2 7.8 66.0 8.1 67.1 8.3 70.6
8.9
1st Cycle 66.4 9.5 66.2 9.4 67.7 9.7 69.7
10.0
B2-3
Last Cycle 64.1 8.5 64.9 8.7 65.8 9.1 67.8
9.4
[00158] The influence of lateral displacements with identical amplitude
history
(AA-pattern) but different rate (30 and 76 mm/s), on lateral response of the
bearings
is shown in Fig. 24. By way of example, only half of the scragged loops at
amplitudes
100%, 150%, and 200% tr are shown in this figure. As can be seen, at the
higher rate,
the bearings exhibited higher lateral load resisting capacity. This was
consistent with
observations from the rate sensitivity tests at 100% tr displacement
amplitude.
Influence of vertical pressure
[00159] To understand the influence of vertical pressure on lateral response,
cyclic
tests under three different levels of vertical pressure were conducted. The
test
procedure for each bearing was as follows: the bearing was vertically loaded
to the
target vertical pressure, next three cycles of fully reversed sinusoidal
lateral
displacements were applied with a frequency of 1 Hz (implying an average
displacement rate of 76 mm/s) and amplitude of 100% tr (19 mm), and finally
the
bearing was vertically unloaded. The three prescribed vertical pressures were,
0.8,
1.6, and 2.4 MPa. The upper and lower limits of the applied vertical pressure
represent 50% variation with respect to the design pressure of 1.6 MPa.
47
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[00160] The influence of vertical pressure on effective horizontal stiffness
and
damping ratio of the bearings can be seen in Fig. 20. Test results showed that
both
effective horizontal stiffness and effective damping increased with the
increased
vertical pressure. However, under the investigated vertical pressure range,
effective
damping was found to be more sensitive to the level of vertical pressure than
was the
effective horizontal stiffness.
[00161] Fiber reinforcement consists of individual fibers grouped into strands
and
woven into a fabric. The fabric is embedded into a thin polymer matrix to form
the
fiber reinforcement sheet. Vertical compression of the bearing causes lateral
expansion in the bearing and corresponding tension in the fabric which tends
to
straighten it out. In this regard, the fiber fabric is less rigid in extension
than the
individual strands which, in turn have lower axial stiffness than the
individual fibers.
The extent of straightening of the fabric, the strands in the fabric and the
fibers in the
strand are affected by the level of vertical compression on the bearing. When
a SU-
FREI bearing is deformed under lateral load, the reinforcement sheets become
curved
due to the lack of flexural rigidity. Therefore, the internal tension in the
curved
reinforcement sheet forces the fibers to slip against each other in the
strands, and
strands to slip against each other in the fabric. Both actions dissipate
energy through
friction. Thus, the additional damping at higher vertical pressure is
attributed to
increased internal tension in the fibers, and higher frictional forces during
interfacial
slippage.
[00162] In conventional steel reinforced elastomeric bearings, the steel
reinforcement layers are relatively rigid in both extension and flexure.
Therefore, they
have little influence on the lateral response of the bearing, which is
primarily
governed by the mechanical properties of the elastomer. However, in FREI
bearings,
due to tensile extensibility and lack of bending rigidity in fiber fabrics,
properties of
the reinforcement are involved in the lateral response of the bearing. This
involvement is much more pronounced for SU-FREI bearings that exhibit rollover
deformation and introduces large curvature in the fabric. Accordingly, any
variations
in the mechanical properties of the fiber fabric reinforcement affect the
lateral
response. Results of tensile tests conducted on coupon specimens made from the
48
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
carbon-reinforcement sheets, showed that up to a certain limit, the effective
tensile
modulus of fiber reinforcement increased with increasing tensile load. It was
found
that an increase in the effective horizontal stiffness of the bearings, with
increased
vertical pressure, was related to the increased effective tensile modulus of
the fiber
reinforcement. It is understood that the pattern of effective horizontal
stiffness
increasing with vertical pressure is not expected to continue at high levels
of vertical
pressure. At some high level of vertical pressure, straightening of the
strands in the
fabric and the fibers in the strands would have been nearly complete so that
any
change in the fiber modulus becomes negligible. In such a case, a descending
pattern
of effective horizontal stiffness with the increasing vertical pressure is
expected and
the vertical pressure approaches the buckling pressure of the bearings.
Similar to an
ordinary column, a multilayered elastomeric bearing is susceptible to buckling
type of
instability. However, the buckling is dominated by low horizontal stiffness of
the
bearing.
[00163] As seen in Fig. 20, the maximum variations in the effective horizontal
stiffness and effective damping were approximately 2% and 10%, respectively.
Similar cyclic tests at 100% tr and 150% tr lateral displacement amplitudes
with the
lower average displacement rate of 30 mm/s were conducted on Bearing B1-1.
Under
the prescribed vertical pressure levels, the bearing remained stable and a
consistent
pattern of increasing horizontal stiffness and damping with increasing
vertical
pressure was observed. The largest variations occurred at 150% tr displacement
amplitude where, compared to tests at 1.6 MPa vertical pressure, the effective
horizontal stiffness decreased by 4% under 1.1 MPa vertical pressure. Under
2.4 MPa
vertical pressure, a maximum 10% increase in the effective damping was
observed
compared to results under 1.6 MPa vertical design pressure.
Vertical compression testing
[00164] The vertical frequency of an elastomeric bearing is an important
design
parameter which is directly controlled by the vertical compressive stiffness
of the
bearing. Vertical testing was conducted to evaluate the corresponding
effective
vertical stiffness of the bearings under the prescribed levels of vertical
pressure.
Vertical tests were conducted in a load control manner during which the
bearing was
49
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
monotonically loaded to the target vertical pressure "r. Next, three fully
reversed
cycles with amplitudes of 20% "r were applied and finally the bearing was
monotonically unloaded (see Fig. 21). In order to study the influence of
vertical
pressure on the vertical stiffness, all of the bearings were tested under
three specific
levels of vertical pressure, namely, 0.82, 1.63, and 2.45 MPa.
[00165] Vertical testing was initially conducted on the virgin Bearings B1-3
and
B2-3. As a representative example, the resulting compressive stress-strain
curves for
Bearing B2-3 at different levels of the vertical pressure are presented in
Fig. 22a. The
slope of a straight line fitted to the cyclic portion of the curve reflects
the effective
compressive modulus of the bearing under the specified pressure. Fig. 22b
shows that
the effective compressive modulus of Bearing B2-3 increased with increasing
vertical
pressure. The same response behavior was observed for all other bearings
tested in
this study. Table 8 contains test results for the bearings mentioned above.
Table 8. Results of initial vertical test on Bearings B1-3 and B2-3
before lateral cyclic testing
Vertical Vertical Vertical Compressive Vertical
Bearing Load Pressure Stiffness Modulus
Frequency
(kN) (MPa) (kN/mm) (MPa) (Hz)
4 0.82 57.5 223.6 59.8
B1-3 8 1.63 88.1 342.4 52.3
12 2.45 111.7 434.3 48.1
4 0.82 59.9 232.8 61.0
B2-3 8 1.63 98.0 380.8 55.2
12 2.45 115.0 447.1 48.8
[00166] As with conventional steel reinforced elastomeric bearings, the
resulting
vertical compressive stress-strain curves of the tested fiber reinforced
bearings are
highly nonlinear with a significant run-in before the full vertical stiffness
is
developed. This nonlinear response is primarily due to the behavior of the
elastomer
in compression. The initial run-in depends on several factors including:
issues related
to workmanship, alignment of the reinforcing layers and, in the case of fiber
reinforced bearings, flexibility of the reinforcement in extension. To reduce
initial
run-in, a special device was designed to apply orthogonal in-plane tension to
the bi-
directional fabric. Using this setup, the carbon fiber fabric utilized in the
construction
of the tested bearings was initially (i.e., before bonding to the elastomer
layers)
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
stretched in two perpendicular directions. While stretched in tension, an
elastomer
layer was bonded to one face of the fabric. The fabric maintained in tension
until the
curing of the bonding compound was completed. Comparing the resulting vertical
load-deflection in this example with those of Example 1 indicates that
increasing the
initial tautness of the fiber reinforcement significantly reduces the
initial run-in in the
load-deflection response. In general, this initial run-in was determined to be
of little
concern as long as the bearings develop an adequate level of vertical
stiffness under
the minimum expected value of the vertical load.
[00167] After completing the lateral cyclic loading tests on all of the
bearings (see
Table 2), vertical testing was performed on Bearings B1-1 and B2-1 and was
repeated
on Bearings B1-3 and B2-3. The main objective was to examine the vertical
performance of bearings that had already been exposed to extreme lateral
displacements. Except for Bearing B1-1, no damage was visible on the other
tested
bearings. Table 9 summarizes the test outcomes where, compared to results from
tests
prior to lateral loading compiled in Table 8, maximum 10% and 12% decreases in
the
vertical stiffness of the bearings at 1.6 MPa and 2.4 MPa vertical pressure,
respectively, were observed. These small changes document acceptably small
damage
to the bearings where, after a major earthquake event, no serious
serviceability issues
would exist regarding the vertical performance of the bearings. After
completion of
the vertical tests, no debonding in the bearings was observed through visual
inspection.
Table 9. Results of vertical test on Bearings B1-1, B1-3, B2-1, and B2-3
after lateral cyclic testing
Vertical Vertical Vertical Compressive Vertical
Bearing Load Pressure Stiffness Modulus
Frequency
(kN) (MPa) (kNimm) (MPa) (Hz)
4 0.82 50.0 194.4 55.7
B1-1* 8 1.63 72.2 280.7 47.4
12 2.45 85.9 334.0 42.2
4 0.82 56.2 218.5 59.1
B1-3 8 1.63 79.5 309.1 49.7
12 2.45 98.2 381.8 45.1
4 0.82 53.9 209.6 57.9
B2-1 8 1.63 81.7 317.6 50.4
12 2.45 105.3 409.4 46.7
4 0.82 58.0 225.5 60.0
B2-3 8 1.63 95.6 371.7 54.5
12 2.45 112.3 436.6 48.2
51
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
* Bearing B1-1 showed the lowest stiffness due to partial debonding in one of
its
laminates after completion of lateral cyclic test at 45 .
[00168] The vertical stiffness of elastomeric bearings is usually several
hundred
times larger than the effective horizontal stiffness of the bearing. This is
to ensure that
rocking vibration modes are prevented from participating in the response of
the base
isolated building. According to Tables 3 and 4, the bearings exhibited
vertical
frequencies ranging from 42 Hz to 64 Hz, depending on the level of the
vertical
compression. These values are significantly higher than the target lateral
isolated
frequency range of the 1/4 scale bearings (2 Hz or less). Accordingly, the
vertical
performance of the bearings is considered quite satisfactory.
Repeatability of the lateral response
[00169] After completion of the entire test program assigned for each bearing
(see
Table 2), lateral cyclic testing, with the displacement history shown in Fig.
14a, was
repeated for all of the bearings. The main objective was to evaluate the
overall lateral
response characteristics and the mechanical properties of the bearings that
were
previously subjected to large lateral displacements at 0 , 90 , and 45
directions. All
bearings were retested at 0 orientation and Bearing B1-3 was also tested at
90 . Fig.
23 shows the lateral response properties versus amplitudes of lateral
displacement.
The observed trends were similar to those of the virgin bearings (see Fig.
19a). Close
agreement in the response of the tested bearings once again suggests that any
variation in the virgin response of the bearings is primarily related to the
unscragged
configuration of the rubber. Once the bearings are scragged, variation in the
lateral
response becomes negligible.
[00170] As a typical plot, hysteresis loops corresponding to all cycles of the
test on
Bearing B1-3 (at 0 orientation) are shown in Fig. 23a. Unlike the previous
set of
cyclic tests, no significant differences were observed in the responses during
the first
cycle and the subsequent cycles at each level of displacement amplitude. This
is not
surprising since the bearing was already scragged. However, as opposed to the
behavior shown in Fig. 14b for unscragged bearings, no significant stiffening
occurred at the extreme portion of lateral cycles except at the largest
lateral
52
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
displacement amplitude of 200% tr which was also the largest previously
applied
displacement. This can be seen in Fig. 25a, where, the scragged half-cycle
loops of
Bearing B1-3 (at 00 orientation) are plotted along with the corresponding
scragged
response of the same bearing at its virgin configuration (Bearing B1-3, at 00
in Fig.
14b). It is postulated that the large lateral displacements in the 900 and 450
directions
resulted in more breakage in the weaker crosslinks of the elastomer and, in
turn, this
influenced the lateral response at 00 orientation in the repeated cyclic
tests. As shown
in Fig. 25b, retested Bearing B1-3 showed lower effective horizontal stiffness
and
higher effective damping relative to its virgin response. At 150% tr lateral
displacement, the maximum variation in the average response was observed with
22%
decrease in stiffness and less than 13% increase in the effective damping.
Despite the
stiffness degradation, all of the bearings remained stable with no negative or
zero
instantaneous tangent stiffness in the loading branch of the hysteresis loops
(see Fig.
24). As such, the lateral performance of the bearings, even after previously
experiencing large lateral displacements at 00, 900, and 450 orientations, was
deemed to be quite satisfactory.
[00171] No significant damage was observed after completing the initial set of
cyclic testing (except for Bearing B1-1 that acquired partial debonding in one
of its
laminates during the diagonal cyclic testing). After completing the repeated
cyclic
testing, no new damage was visible in Bearing B1-1, and no debonding in the
other
bearings was visible. Additionally, consistent response of the bearings and
the
repeatability of the hysteresis loops during cycles with constant amplitude
suggested
that if any unobserved internal delamination had occurred in the bearings, it
had
negligible influence on bearing behavior. Thereby, the resulting softening in
the
bearings' response, compared to their original response, is thought to have
occurred
from the Mullin's effect in the elastomer. Mullin's effect attributes to
strength
degradation feature of an elastomer compound in the subsequent cycles of
loading
with respect to the first load cycle. Since no vertical tension is developed
in a laterally
deformed SU-FREI bearing, internal bonding between elastomer and reinforcement
sheets are subjected to a much lower demand compared to the same bearing in a
bonded application. Accordingly, a SU-FREI is able to survive multiple cycles
of
large amplitude lateral displacements.
53
CA 02727039 2014-09-26
[00172] A close examination of Figs. 19b and 23a shows that, except at the
largest
displacement amplitude (200% tr), Bearings B1-3, B2-1, and B2-3 exhibited
larger
effective horizontal stiffness in the repeated cyclic testing than in the
cyclic testing
with descending amplitudes. This is another indication that no significant
damage was
done to these bearings even when the repeated cyclic tests were performed.
Since both
Figs. 19b and 23a present the scragged properties of the bearings, the lower
stiffness
values in Fig. 19b imply that there is an influence of lateral displacement
history on
the bearings response even if they were already scragged. It appears that
cyclic lateral
displacements with a descending amplitude pattern produce lower bound results
for
the effective horizontal stiffness and perhaps an acceptable estimate of the
upper
bound for the effective damping of the bearings.
Extreme lateral response
[00173] The full contact vertical face lateral displacement (50 is defined as
the
point when the originally vertical faces of the bearings completely contact
the upper
and lower platens of the test machine. The lateral displacement capacity
(6max) is
reached when, due to rollout instability, the bearing's lateral load
displacement
hysteresis behavior exhibits softening behavior with negative tangential
stiffness. For
SU-FREI bearings that exhibit SR-deformation in an unbonded application, S.),
is
significantly larger than 6fe provided that no significant damage occurs in
the bearing.
In order to acquire insight into the lateral displacements 6f, and 8max and
the extreme
lateral response behavior of the bearings, Bearing TB1 was constructed with
the same
characteristics of the previous bearings but with a different width. This
bearing had a
70 x 41 mm physical dimensions in plan, a total thickness of approximately 25
mm
and total thickness of the Neoprene layers t, = 19 mm. This bearing had an
aspect
ratio of R = 2.8 along its longitudinal direction. The shape factor for the
bearing, due
to a reduced width, was S = 8.2. Unlike all of the previous bearings, no
finishing coats
of the bonding compound were applied to the perimeter surfaces of this bearing
as
there was an interest in studying the role of finishing coats on damage
protection of
the bearing at large lateral displacements.
54
CA 02727039 2014-09-26
[00174] Fig. 26 contains the lateral load-displacement hysteresis loops for
all cycles
of the cyclic test on the virgin Bearing TB1 laterally loaded in its
longitudinal
direction. Under a constant 1.6 MPa vertical pressure, the bearing was
subjected to the
displacement history of Fig. 14a. As seen in Fig. 26, the bearing exhibited an
acceptable SR-deformation. No delamination or any other damage was visible in
the
bearing after completion of this cyclic testing. This implied that lateral
displacements
up to 200% t, were safely accommodated by the bearing even in the absence of
the
finishing coats. Observations from all cyclic testing previously conducted in
this study
implied that 8k was larger than the 200% t, lateral displacement amplitude. As
shown
in Fig. 27b, the original vertical faces of the bearings did not completely
meet the
upper and lower platens at the 200% tr lateral displacement.
[00175] To study the ultimate response of the bearing, the displacement time
history
shown in Fig. 28a with four displacement amplitudes, was applied to the
bearing
while a vertical pressure of 1.6 MPa was maintained. Having been previously
scragged, no significant unscragged properties were observed up to amplitude
of
200% tr. However, during the first cycle of each 250% and 300% tr,
displacement
amplitude, the bearing showed larger effective horizontal stiffness and
damping
during than the following cycles. This is considered to be due to Mullin's
effect.
[00176] Fig. 28b shows the hysteresis loops for the last cycle of each
displacement
amplitude where positive incremental lateral load-resisting capacity versus
displacement was observed for the 250% tr (47.5 mm) displacement amplitude.
This
displacement amplitude corresponds to .3r, (see Fig. 27c). The bearing showed
acceptable SR-deformation during the first 2 cycles of 300% t, (57 mm)
amplitude.
However, the hysteresis loop for the final cycle of the test showed
significant stiffness
degradation as a result of a major delamination localized in one of the
bearing's
laminates (see Fig. 270. As seen in Fig. 28b, the hysteresis loop at the
largest lateral
displacement exhibited approximately zero tangent stiffness (at positive
amplitudes)
and even some softening behavior (at negative amplitudes) which were followed
by a
stiffening behavior at the largest level of lateral displacement. The
stiffening behavior
at the extreme portion of 300% t, lateral displacement cycle implied that the
line of
action of the resultant vertical load had not reached the edge of the
bearing's contact
CA 02727039 2014-09-26
surface. Accordingly, if the localized delamination in the bearing had not
occurred,
the bearing would have remained stable for lateral displacements exceeding
300% tr.
Despite this, for Bearing TB1 8 is conservatively considered to be as low
as 6f, =
250% tr. As shown in Fig. 27f, at the end of the last test cycle, the bearing
had slipped
approximately 6 mm in the direction of the input displacement, with respect to
its
original position. This is 10.5% of the 300% tr lateral displacement
amplitude.
[00177] Testing on Bearing TB1 provided insight into the extreme lateral
response
of such bearings. Unlike Bearing TB1, the lateral faces of all other bearings
were
covered with 2 coats of the bonding compound. Therefore, based on the above
test
results, these bearings are expected to safely sustain a 250% tr (47.5 mm)
lateral
displacement amplitude. For a larger lateral displacement demand, the
bearing's
design specifications could be consistently altered in order to meet the
desired
maximum lateral displacement. In some practical cases, even if the bearings'
lateral
displacement capacity meets the extreme design expectations, special
protectors may
be utilized to prevent the superstructure from experiencing any unanticipated
excessive displacements at the bearings. This would depend on the importance
of the
structure and is a precaution employed in application of conventional base
isolators.
Summary of Example 3
[00178] Four 1/4 scale square FRET bearings were experimentally studied under
various test sequences in this research program. The test bearings were
employed in
an unbonded application. The test results showed that this application
maximized the
efficiency of the bearings as seismic isolators. Response characteristics of
the tested
SU-FRE1 bearings were found to be similar to those of conventional steel
reinforced
high-damped rubber bearings in the sense of changes in the effective
horizontal
stiffness and damping with change in the amplitude of displacement even though
an
unfilled elastomer was utilized.
[00179] In general, the lateral response of the SU-FREI bearings was found to
be
highly nonlinear with some sensitivity to the amplitude, amplitude history and
rate of
the cyclic lateral displacements imposed to the bearings, and to the level of
vertical
56
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
pressure on the bearings. Additionally, the transition between virgin and
scragged
properties of the elastomer, adds to the complexity of the response.
[00180] Lateral displacement amplitude: Advantageously, the effective
horizontal
stiffness of the bearings decreased significantly (approximately 60%) with
increasing
amplitude of lateral displacements. The equivalent viscous damping ratios of
the
bearings also decreased with increasing lateral displacement amplitude.
However,
sufficient damping (minimum 8%) was achieved at the extreme lateral
displacements.
[00181] Lateral displacement amplitude history: The shape of hysteresis loops
was
highly influenced by the history of the input lateral displacements. The
virgin
bearings that were subjected to displacement histories with an ascending
pattern of
amplitude exhibited unscragged response properties during the first cycle of
each
displacement amplitude. The maximum differences between the scragged and
unscragged response values were observed at amplitudes larger than 150% tr.
The
unscragged effective horizontal stiffness and damping values were at most 10%
and
19%, respectively, larger than the average response value of the scragged
(stable)
loops (virgin Bearing B1-3, 0 in Table 7). The minimum values of the
effective
horizontal stiffness and the maximum effective damping ratio occurred when the
bearings were subjected to a descending pattern of displacement amplitude.
Such a
lateral displacement history resulted in a maximum of 27% decrease in
effective
horizontal stiffness and 15% increase in the effective damping compare to the
average
response of the corresponding virgin bearing against an ascending amplitude of
lateral
displacement history. This is shown when comparing the responses of the virgin
Bearing B1-3, 0 and Bearing B1-3, 90 at 150% tr in Table 7.
[00182] Lateral displacement rate: Lateral response properties were found to
be
increased with increasing displacement rate. Test results showed that
effective
damping was more sensitive to the lateral displacement rate than was the
effective
horizontal stiffness. However, since the bearings were tested at an adequate
average
displacement rate which was sufficiently close to the actual operating average
rate of
the bearings, the influence of rate on the obtained response values can be
neglected.
57
CA 02727039 2014-09-26
[00183] Vertical pressure: The influence of 50% variation in the 1.6 MPa
vertical
design pressure on the effective horizontal stiffness was found to be
negligible. This
implied that the prescribed values of vertical pressure were sufficiently
below the
buckling load of the bearings. However, the effective damping was found to be
more
influenced by variations in the vertical pressure. A 50% increase in the
vertical
pressure resulted in a 10% increase in the effective damping. For many
practical cases
this variation can be neglected from an engineering perspective.
[00184] For many practical applications, by neglecting the influences of
lateral
displacement rate and the vertical pressure on the lateral response of the
bearings, the
hysteresis loops corresponding to the cyclic testing with ascending and
descending
amplitude patterns indicate the bounds of the bearing's lateral response.
[00185] Vertical compression testing revealed that the SU-FREI bearings tested
in
this example had an acceptable vertical stiffness. Depending on the level of
vertical
load, the resulting vertical frequencies ranged from 42 to 64 Hz. The minimum
achieved vertical frequency of around 40 Hz in the 1/4 scale bearings
represents a
value of 20 Hz in the corresponding full scale bearings. This value is
sufficient as it is
significantly larger than the target isolated frequency of the full scale SU-
FREI
bearings (i.e., 0.63 to 1.25 Hz). After conducting the lateral test program,
subsequent
lateral tests showed that the vertical stiffness degradation of the bearings
was
negligible.
[00186] At 6fc = 250% tr (47.5 mm), the originally vertical faces of the
bearings
were expected to completely contact the upper and lower supports. For SU-FREI
bearings that remain undamaged under large lateral displacements, the lateral
displacement at rollout instability (lateral displacement capacity max) [ 1 11
is expected
to be significantly larger than 6fe. All of the bearings showed SR deformation
during
the original cyclic testing in the 0 , 90 , and 45 directions and during the
repeated
cyclic test in the 0 direction. Consistent lateral response of the scragged
bearings
implied that any significant discrepancy in the virgin response of the
bearings resulted
from unscragged properties of the elastomer. Cyclic testing did not result in
any
58
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
significant damage to the bearings (except Bearing B1-1). This is evidence of
the
reliability of the bearings.
[00187] Considering the lower and upper limits of the calculated base isolated
periods for the tested 'A scale bearings, the isolated periods for the
corresponding full
scale model of the bearings would range from 0.8 to 1.4 s in the
parallel/perpendicular
direction, and 1 to 1.6 s in the diagonal direction. Additionally, the
corresponding
lateral displacement capacity would be 190 mm (250% tr) in conjunction with
sufficient damping (8% and higher). Therefore, the corresponding full scale
bearings
can be effectively used for hazard mitigation of many structures, including
but not
limited to, low-rise buildings, with fixed base period of 0.1 to 0.2 s,
located in
moderate to high seismic regions worldwide.
Alteration of Shapes
[00188] Stable rollover is achieved by proper dimensioning of isolator
bearings
(200) supported within flat contact surfaces (205) and (210), as shown in Fig.
29a.
Here, the vertical faces (215) have a straight edge, and are perpendicular to
the
contact surfaces (205) and (210). Other shapes of bearings are contemplated in
the
present invention. As shown in Fig. 29b,the vertical faces (225) of the
bearing (220)
can be shaped to improve the bearing's lateral response, and accelerate,
delay, or
modify the stiffening effect at the extreme stage of lateral deformation
following the
stable rollover behaviour. Similarly, as shown in Fig. 29c, the vertical faces
(235) of
the bearing (230) are shaped to perform the same function as in Fig. 29b. In
addition,
as shown in Figs. 29d and 29e, the surrounding contact surfaces ((240) and
(245)
above and below the rectangular bearing (200) can be shaped to improve the
bearing's
lateral response and to accelerate, delay, or modify the stiffening effect at
the extreme
stage of lateral deformation following the stable rollover behaviour. Finally,
Figs. 29f
and 29g show how the vertical faces of the rectangular bearing (200) can be
fitted
with preformed shapes (260) and (265) to improve the bearing's lateral
response and
to accelerate or delay the stiffening effect at the extreme stage of lateral
deformation.
There are numerous possible shapes that can be used depending on specific
criteria
related to changing the fundamental period of the bearing, changing damping,
and
"designing-in" the type of stiffening required for stable rollover.
59
CA 02727039 2010-12-06
WO 2008/148203
PCT/CA2008/001077
[00189] The foregoing has constituted a description of specific embodiments
showing how the invention may be applied and put into use. These embodiments
are
only exemplary. The invention in its broadest, and more specific aspects, is
further
described and defined in the claims which now follow.
[00190] These claims, and the language used therein, are to be understood in
terms
of the variants of the invention which have been described. They are not to be
restricted to such variants, but are to be read as covering the full scope of
the
invention as is implicit within the invention and the disclosure that has been
provided
herein.
